notebook_gli_imBruttiti.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Statistics Under the Stars Hackathon (SUS7)\n",
    "This is the solution produced by the team \"The ImBrutiti\" for the Statistics Under the Stars (SUS7) hackathon, organized by y-SIS (Statistical Society of Italy) by the University of Salento at Lecce. Italy and sponsored by Adecco.  \n",
    "\n",
    "Our team, composed of Gianluca Procopio, Giorgio Bertone, Filippo Parlanti and Himel Ghosh, reached the 2nd place and received the \"Best Objective Prediction\" award."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import MultiLabelBinarizer, StandardScaler\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "from geopy.distance import geodesic\n",
    "from geopy.geocoders import Nominatim\n",
    "import time\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from xgboost import XGBClassifier\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from sklearn.metrics import accuracy_score\n",
    "from math import radians, sin, cos, sqrt, atan2\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import job position datasets\n",
    "job_positions_train=pd.read_csv(\"job_positions_train.csv\")\n",
    "job_positions_test=pd.read_csv(\"job_positions_test.csv\")\n",
    "# Import candidates datasets\n",
    "candidates_train=pd.read_csv(\"candidates_train.csv\")\n",
    "candidates_test=pd.read_csv(\"candidates_test.csv\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quick look to how the dataset we are dealing with looks like."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>Prof_cat</th>\n",
       "      <th>Hard_skills_group</th>\n",
       "      <th>Soft_skills</th>\n",
       "      <th>Education</th>\n",
       "      <th>Languages</th>\n",
       "      <th>Cards</th>\n",
       "      <th>Work_city</th>\n",
       "      <th>Latitudine</th>\n",
       "      <th>Longitudine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1739_114</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>Grande distribuzione</td>\n",
       "      <td>Orientamento al cliente;Orientamento al risultato</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.5050</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1606_139</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>Grande distribuzione</td>\n",
       "      <td>Pianificazione e organizzazione</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B</td>\n",
       "      <td>MILANO</td>\n",
       "      <td>45.4773</td>\n",
       "      <td>9.18150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1636_143</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>Contact Center;Vendita;Segreteria</td>\n",
       "      <td>Decision Making;Team Working;Gestione dei camb...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>MILANO</td>\n",
       "      <td>45.4773</td>\n",
       "      <td>9.18150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1739_397</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>Vendita</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.5050</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1739_405</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>Vendita</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.5050</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2526</th>\n",
       "      <td>0264_24098</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>Acquisti;Fogli di calcolo / elettronici;Word P...</td>\n",
       "      <td>Problem solving e analisi;Comunicazione;Team W...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Inglese</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BARASSO</td>\n",
       "      <td>45.8405</td>\n",
       "      <td>8.75683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2527</th>\n",
       "      <td>0300_24353</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>Word Processor - Elaborazione testi;Vendita;Ba...</td>\n",
       "      <td>Decision Making;Problem solving e analisi;Orie...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Inglese</td>\n",
       "      <td>B</td>\n",
       "      <td>SANTA MARIA A MONTE</td>\n",
       "      <td>43.6999</td>\n",
       "      <td>10.68900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2528</th>\n",
       "      <td>0044_24693</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>Segreteria;Fogli di calcolo / elettronici</td>\n",
       "      <td>Problem solving e analisi;Orientamento alla qu...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B</td>\n",
       "      <td>CAORSO</td>\n",
       "      <td>45.0485</td>\n",
       "      <td>9.87107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2529</th>\n",
       "      <td>0016_27490</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>Fogli di calcolo / elettronici;Animazione e sp...</td>\n",
       "      <td>Comunicazione;Decision Making;Pianificazione e...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Inglese</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRENTO</td>\n",
       "      <td>46.0679</td>\n",
       "      <td>11.12110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2530</th>\n",
       "      <td>0154_37589</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>Vendita;Segreteria;Fogli di calcolo / elettron...</td>\n",
       "      <td>Team Working;Decision Making;Orientamento alla...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Inglese</td>\n",
       "      <td>B</td>\n",
       "      <td>CIRIE'</td>\n",
       "      <td>45.2380</td>\n",
       "      <td>7.57680</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2531 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          Job_ID                                    Prof_cat  \\\n",
       "0       1739_114  addetto agli scaffali/addetta agli scafali   \n",
       "1       1606_139  addetto agli scaffali/addetta agli scafali   \n",
       "2       1636_143  addetto agli scaffali/addetta agli scafali   \n",
       "3       1739_397  addetto agli scaffali/addetta agli scafali   \n",
       "4       1739_405  addetto agli scaffali/addetta agli scafali   \n",
       "...          ...                                         ...   \n",
       "2526  0264_24098                       segretario/segretaria   \n",
       "2527  0300_24353                       segretario/segretaria   \n",
       "2528  0044_24693                       segretario/segretaria   \n",
       "2529  0016_27490                       segretario/segretaria   \n",
       "2530  0154_37589                       segretario/segretaria   \n",
       "\n",
       "                                      Hard_skills_group  \\\n",
       "0                                  Grande distribuzione   \n",
       "1                                  Grande distribuzione   \n",
       "2                     Contact Center;Vendita;Segreteria   \n",
       "3                                               Vendita   \n",
       "4                                               Vendita   \n",
       "...                                                 ...   \n",
       "2526  Acquisti;Fogli di calcolo / elettronici;Word P...   \n",
       "2527  Word Processor - Elaborazione testi;Vendita;Ba...   \n",
       "2528          Segreteria;Fogli di calcolo / elettronici   \n",
       "2529  Fogli di calcolo / elettronici;Animazione e sp...   \n",
       "2530  Vendita;Segreteria;Fogli di calcolo / elettron...   \n",
       "\n",
       "                                            Soft_skills Education Languages  \\\n",
       "0     Orientamento al cliente;Orientamento al risultato       NaN       NaN   \n",
       "1                       Pianificazione e organizzazione       NaN       NaN   \n",
       "2     Decision Making;Team Working;Gestione dei camb...       NaN       NaN   \n",
       "3                                                   NaN       NaN       NaN   \n",
       "4                                                   NaN       NaN       NaN   \n",
       "...                                                 ...       ...       ...   \n",
       "2526  Problem solving e analisi;Comunicazione;Team W...       NaN   Inglese   \n",
       "2527  Decision Making;Problem solving e analisi;Orie...       NaN   Inglese   \n",
       "2528  Problem solving e analisi;Orientamento alla qu...       NaN       NaN   \n",
       "2529  Comunicazione;Decision Making;Pianificazione e...       NaN   Inglese   \n",
       "2530  Team Working;Decision Making;Orientamento alla...       NaN   Inglese   \n",
       "\n",
       "     Cards            Work_city  Latitudine  Longitudine  \n",
       "0      NaN            PIOLTELLO     45.5050      9.33151  \n",
       "1        B               MILANO     45.4773      9.18150  \n",
       "2      NaN               MILANO     45.4773      9.18150  \n",
       "3        B            PIOLTELLO     45.5050      9.33151  \n",
       "4        B            PIOLTELLO     45.5050      9.33151  \n",
       "...    ...                  ...         ...          ...  \n",
       "2526   NaN              BARASSO     45.8405      8.75683  \n",
       "2527     B  SANTA MARIA A MONTE     43.6999     10.68900  \n",
       "2528     B               CAORSO     45.0485      9.87107  \n",
       "2529   NaN               TRENTO     46.0679     11.12110  \n",
       "2530     B               CIRIE'     45.2380      7.57680  \n",
       "\n",
       "[2531 rows x 10 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "job_positions_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since attributes like \"Soft Skill\", \"Hard Skill\" and \"Prof_cat\" contain a list of string separated by semicolon, we need to represent them in another way.   \n",
    "Therefore we first transformed them into a set of string by splitting on semicolons."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert lists into sets for easier processing\n",
    "def convert_to_set(column):\n",
    "    return column.apply(lambda x: set(x.split(';')) if x else set())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DATASET JOB_POSITION\n",
    "# Train\n",
    "# remove NaN field\n",
    "job_positions_train.fillna('', inplace=True)\n",
    "job_positions_train['Hard_skills_group'] = convert_to_set(job_positions_train['Hard_skills_group'])\n",
    "job_positions_train['Soft_skills'] = convert_to_set(job_positions_train['Soft_skills'])\n",
    "job_positions_train['Languages'] = convert_to_set(job_positions_train['Languages'])\n",
    "job_positions_train['Cards'] = convert_to_set(job_positions_train['Cards'])\n",
    "job_positions_train['Education'] = convert_to_set(job_positions_train['Education'])\n",
    "\n",
    "# Test\n",
    "# remove NaN field\n",
    "job_positions_test.fillna('', inplace=True)\n",
    "job_positions_test['Hard_skills_group'] = convert_to_set(job_positions_test['Hard_skills_group'])\n",
    "job_positions_test['Soft_skills'] = convert_to_set(job_positions_test['Soft_skills'])\n",
    "job_positions_test['Languages'] = convert_to_set(job_positions_test['Languages'])\n",
    "job_positions_test['Cards'] = convert_to_set(job_positions_test['Cards'])\n",
    "job_positions_test['Education'] = convert_to_set(job_positions_test['Education'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>Prof_cat</th>\n",
       "      <th>Hard_skills_group</th>\n",
       "      <th>Soft_skills</th>\n",
       "      <th>Education</th>\n",
       "      <th>Languages</th>\n",
       "      <th>Cards</th>\n",
       "      <th>Work_city</th>\n",
       "      <th>Latitudine</th>\n",
       "      <th>Longitudine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1739_114</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>{Grande distribuzione}</td>\n",
       "      <td>{Orientamento al cliente, Orientamento al risu...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.505</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1606_139</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>{Grande distribuzione}</td>\n",
       "      <td>{Pianificazione e organizzazione}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>MILANO</td>\n",
       "      <td>45.4773</td>\n",
       "      <td>9.1815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1636_143</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>{Vendita, Segreteria, Contact Center}</td>\n",
       "      <td>{Gestione dei cambiamenti, Decision Making, Te...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>MILANO</td>\n",
       "      <td>45.4773</td>\n",
       "      <td>9.1815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1739_397</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>{Vendita}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.505</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1739_405</td>\n",
       "      <td>addetto agli scaffali/addetta agli scafali</td>\n",
       "      <td>{Vendita}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>PIOLTELLO</td>\n",
       "      <td>45.505</td>\n",
       "      <td>9.33151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2526</th>\n",
       "      <td>0264_24098</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>{Word Processor - Elaborazione testi, Vendita,...</td>\n",
       "      <td>{Orientamento alla qualità, Problem solving e ...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{Inglese}</td>\n",
       "      <td>{}</td>\n",
       "      <td>BARASSO</td>\n",
       "      <td>45.8405</td>\n",
       "      <td>8.75683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2527</th>\n",
       "      <td>0300_24353</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>{Word Processor - Elaborazione testi, Vendita,...</td>\n",
       "      <td>{Problem solving e analisi, Orientamento alla ...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{Inglese}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>SANTA MARIA A MONTE</td>\n",
       "      <td>43.6999</td>\n",
       "      <td>10.689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2528</th>\n",
       "      <td>0044_24693</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>{Segreteria, Fogli di calcolo / elettronici}</td>\n",
       "      <td>{Problem solving e analisi, Orientamento alla ...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>CAORSO</td>\n",
       "      <td>45.0485</td>\n",
       "      <td>9.87107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2529</th>\n",
       "      <td>0016_27490</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>{Word Processor - Elaborazione testi, Progetta...</td>\n",
       "      <td>{Orientamento alla qualità, Problem solving e ...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{Inglese}</td>\n",
       "      <td>{}</td>\n",
       "      <td>TRENTO</td>\n",
       "      <td>46.0679</td>\n",
       "      <td>11.1211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2530</th>\n",
       "      <td>0154_37589</td>\n",
       "      <td>segretario/segretaria</td>\n",
       "      <td>{Word Processor - Elaborazione testi, Vendita,...</td>\n",
       "      <td>{Problem solving e analisi, Orientamento alla ...</td>\n",
       "      <td>{}</td>\n",
       "      <td>{Inglese}</td>\n",
       "      <td>{B}</td>\n",
       "      <td>CIRIE'</td>\n",
       "      <td>45.238</td>\n",
       "      <td>7.5768</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2531 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          Job_ID                                    Prof_cat  \\\n",
       "0       1739_114  addetto agli scaffali/addetta agli scafali   \n",
       "1       1606_139  addetto agli scaffali/addetta agli scafali   \n",
       "2       1636_143  addetto agli scaffali/addetta agli scafali   \n",
       "3       1739_397  addetto agli scaffali/addetta agli scafali   \n",
       "4       1739_405  addetto agli scaffali/addetta agli scafali   \n",
       "...          ...                                         ...   \n",
       "2526  0264_24098                       segretario/segretaria   \n",
       "2527  0300_24353                       segretario/segretaria   \n",
       "2528  0044_24693                       segretario/segretaria   \n",
       "2529  0016_27490                       segretario/segretaria   \n",
       "2530  0154_37589                       segretario/segretaria   \n",
       "\n",
       "                                      Hard_skills_group  \\\n",
       "0                                {Grande distribuzione}   \n",
       "1                                {Grande distribuzione}   \n",
       "2                 {Vendita, Segreteria, Contact Center}   \n",
       "3                                             {Vendita}   \n",
       "4                                             {Vendita}   \n",
       "...                                                 ...   \n",
       "2526  {Word Processor - Elaborazione testi, Vendita,...   \n",
       "2527  {Word Processor - Elaborazione testi, Vendita,...   \n",
       "2528       {Segreteria, Fogli di calcolo / elettronici}   \n",
       "2529  {Word Processor - Elaborazione testi, Progetta...   \n",
       "2530  {Word Processor - Elaborazione testi, Vendita,...   \n",
       "\n",
       "                                            Soft_skills Education  Languages  \\\n",
       "0     {Orientamento al cliente, Orientamento al risu...        {}         {}   \n",
       "1                     {Pianificazione e organizzazione}        {}         {}   \n",
       "2     {Gestione dei cambiamenti, Decision Making, Te...        {}         {}   \n",
       "3                                                    {}        {}         {}   \n",
       "4                                                    {}        {}         {}   \n",
       "...                                                 ...       ...        ...   \n",
       "2526  {Orientamento alla qualità, Problem solving e ...        {}  {Inglese}   \n",
       "2527  {Problem solving e analisi, Orientamento alla ...        {}  {Inglese}   \n",
       "2528  {Problem solving e analisi, Orientamento alla ...        {}         {}   \n",
       "2529  {Orientamento alla qualità, Problem solving e ...        {}  {Inglese}   \n",
       "2530  {Problem solving e analisi, Orientamento alla ...        {}  {Inglese}   \n",
       "\n",
       "     Cards            Work_city Latitudine Longitudine  \n",
       "0       {}            PIOLTELLO     45.505     9.33151  \n",
       "1      {B}               MILANO    45.4773      9.1815  \n",
       "2       {}               MILANO    45.4773      9.1815  \n",
       "3      {B}            PIOLTELLO     45.505     9.33151  \n",
       "4      {B}            PIOLTELLO     45.505     9.33151  \n",
       "...    ...                  ...        ...         ...  \n",
       "2526    {}              BARASSO    45.8405     8.75683  \n",
       "2527   {B}  SANTA MARIA A MONTE    43.6999      10.689  \n",
       "2528   {B}               CAORSO    45.0485     9.87107  \n",
       "2529    {}               TRENTO    46.0679     11.1211  \n",
       "2530   {B}               CIRIE'     45.238      7.5768  \n",
       "\n",
       "[2531 rows x 10 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "job_positions_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DATASET CANDIDATES\n",
    "# Train\n",
    "# Fill missing values with empty strings\n",
    "candidates_train.fillna('', inplace=True)\n",
    "\n",
    "candidates_train['Prof_cat_exp'] = convert_to_set(candidates_train['Prof_cat_exp'])\n",
    "candidates_train['Prof_cat_int'] = convert_to_set(candidates_train['Prof_cat_int'])\n",
    "candidates_train['Hard_skills_group'] = convert_to_set(candidates_train['Hard_skills_group'])\n",
    "candidates_train['Soft_skills'] = convert_to_set(candidates_train['Soft_skills'])\n",
    "candidates_train['Languages'] = convert_to_set(candidates_train['Languages'])\n",
    "candidates_train['Cards'] = convert_to_set(candidates_train['Cards'])\n",
    "candidates_train['Education']=convert_to_set(candidates_train['Education'])\n",
    "\n",
    "# Test\n",
    "# Fill missing values with empty strings\n",
    "candidates_test.fillna('', inplace=True)\n",
    "\n",
    "candidates_test['Prof_cat_exp'] = convert_to_set(candidates_test['Prof_cat_exp'])\n",
    "candidates_test['Prof_cat_int'] = convert_to_set(candidates_test['Prof_cat_int'])\n",
    "candidates_test['Hard_skills_group'] = convert_to_set(candidates_test['Hard_skills_group'])\n",
    "candidates_test['Soft_skills'] = convert_to_set(candidates_test['Soft_skills'])\n",
    "candidates_test['Languages'] = convert_to_set(candidates_test['Languages'])\n",
    "candidates_test['Cards'] = convert_to_set(candidates_test['Cards'])\n",
    "candidates_test['Education']=convert_to_set(candidates_test['Education'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we noticed that the Education column was really bad formatted, therefore we tried to map the Candidate's Education fields to the 4 different categories in the job Education field ('Diploma / Accademia', 'Istituto Tecnico Superiore (ITS)', 'Laurea Magistrale', 'Laurea Triennale', 'Qualifica / Attestato'), but we merged 'Diploma / Accademia' and 'Qualifica / Attestato' into the category 'Other'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Job Education has:\n",
    "# ['Diploma / Accademia', 'Istituto Tecnico Superiore (ITS)', 'Laurea Magistrale', 'Laurea Triennale', 'Qualifica / Attestato']\n",
    "\n",
    "## Education Mapping\n",
    "# Containing word ITS --> 'Istituto Tecnico Superiore (ITS)'\n",
    "# Containing (L-##) --> 'Laurea Triennale'\n",
    "# Containing (LM-##) --> 'Laurea Magistrale'\n",
    "# Other things --> 'Other'\n",
    "def map_education(edu):\n",
    "    edu_set = set(edu)\n",
    "    out = set()\n",
    "    for x in edu_set:\n",
    "        if 'ITS' in x:\n",
    "            out.add('Istituto Tecnico Superiore (ITS)')\n",
    "        elif '(L-' in x:\n",
    "            out.add('Laurea Triennale')\n",
    "        elif '(LM-' in x:\n",
    "            out.add('Laurea Magistrale')\n",
    "        else:\n",
    "            out.add('Other')\n",
    "    return {x for x in out}\n",
    "\n",
    "## Remap 'Diploma / Accademia' and 'Qualifica / Attestato' of Job Education to 'Other'\n",
    "def remap_education_job(edu):\n",
    "    edu_set = set(edu)\n",
    "    out = set()\n",
    "    for x in edu_set:\n",
    "        if 'Diploma / Accademia' in edu:\n",
    "            out.add('Other')\n",
    "        elif 'Qualifica / Attestato' in edu:\n",
    "            out.add('Other')\n",
    "    return {x for x in out}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Education Mapping\n",
    "candidates_test['Education'] = candidates_test['Education'].apply(lambda x: map_education(x))\n",
    "job_positions_train['Education'] = job_positions_train['Education'].apply(lambda x: remap_education_job(x))\n",
    "candidates_train['Education'] = candidates_train['Education'].apply(lambda x: map_education(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To properly deal with 'Hard Skills', 'Soft Skills', 'Languages', 'Cards' and 'Education', once trasformed from string into set, we extracted the combined list of all the single element for each feature and we created a binary matrix by taking advantage of the MultiLabelBinarizer() from the scikitlearn library.  \n",
    "\n",
    "In the end we will get all the features encoded through a binary matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Combine data from both job positions and candidates for fitting\n",
    "all_hard_skills = pd.concat([job_positions_train['Hard_skills_group'], candidates_train['Hard_skills_group']])\n",
    "all_soft_skills = pd.concat([job_positions_train['Soft_skills'], candidates_train['Soft_skills']])\n",
    "all_languages = pd.concat([job_positions_train['Languages'], candidates_train['Languages']])\n",
    "all_cards = pd.concat([job_positions_train['Cards'], candidates_train['Cards']])\n",
    "all_edu = pd.concat([job_positions_train['Education'], candidates_train['Education']])\n",
    "\n",
    "# Combine data from both job positions and candidates for fitting\n",
    "all_hard_skills_t = pd.concat([job_positions_test['Hard_skills_group'], candidates_test['Hard_skills_group']])\n",
    "all_soft_skills_t = pd.concat([job_positions_test['Soft_skills'], candidates_test['Soft_skills']])\n",
    "all_languages_t = pd.concat([job_positions_test['Languages'], candidates_test['Languages']])\n",
    "all_cards_t = pd.concat([job_positions_test['Cards'], candidates_test['Cards']])\n",
    "all_edu_t = pd.concat([job_positions_test['Education'], candidates_test['Education']])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use the MultiLabelBinarizer() to encode the features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize MultiLabelBinarizer\n",
    "mlb = MultiLabelBinarizer()\n",
    "\n",
    "# Fit MultiLabelBinarizer on combined data\n",
    "mlb.fit(all_hard_skills)\n",
    "job_positions_hard_skills = pd.DataFrame(mlb.transform(job_positions_train['Hard_skills_group']), columns=mlb.classes_)\n",
    "candidates_hard_skills = pd.DataFrame(mlb.transform(candidates_train['Hard_skills_group']), columns=mlb.classes_)\n",
    "\n",
    "# Fit MultiLabelBinarizer on combined data\n",
    "job_positions_hard_skills_t = pd.DataFrame(mlb.transform(job_positions_test['Hard_skills_group']), columns=mlb.classes_)\n",
    "candidates_hard_skills_t = pd.DataFrame(mlb.transform(candidates_test['Hard_skills_group']), columns=mlb.classes_)\n",
    "\n",
    "\n",
    "mlb.fit(all_soft_skills)\n",
    "job_positions_soft_skills = pd.DataFrame(mlb.transform(job_positions_train['Soft_skills']), columns=mlb.classes_)\n",
    "candidates_soft_skills = pd.DataFrame(mlb.transform(candidates_train['Soft_skills']), columns=mlb.classes_)\n",
    "\n",
    "job_positions_soft_skills_t = pd.DataFrame(mlb.transform(job_positions_test['Soft_skills']), columns=mlb.classes_)\n",
    "candidates_soft_skills_t = pd.DataFrame(mlb.transform(candidates_test['Soft_skills']), columns=mlb.classes_)\n",
    "\n",
    "\n",
    "mlb.fit(all_languages)\n",
    "job_positions_languages = pd.DataFrame(mlb.transform(job_positions_train['Languages']), columns=mlb.classes_)\n",
    "candidates_languages = pd.DataFrame(mlb.transform(candidates_train['Languages']), columns=mlb.classes_)\n",
    "\n",
    "job_positions_languages_t = pd.DataFrame(mlb.transform(job_positions_test['Languages']), columns=mlb.classes_)\n",
    "candidates_languages_t = pd.DataFrame(mlb.transform(candidates_test['Languages']), columns=mlb.classes_)\n",
    "\n",
    "\n",
    "mlb.fit(all_cards)\n",
    "job_positions_cards = pd.DataFrame(mlb.transform(job_positions_train['Cards']), columns=mlb.classes_)\n",
    "candidates_cards = pd.DataFrame(mlb.transform(candidates_train['Cards']), columns=mlb.classes_)\n",
    "\n",
    "job_positions_cards_t = pd.DataFrame(mlb.transform(job_positions_test['Cards']), columns=mlb.classes_)\n",
    "candidates_cards_t = pd.DataFrame(mlb.transform(candidates_test['Cards']), columns=mlb.classes_)\n",
    "\n",
    "\n",
    "mlb.fit(all_edu)\n",
    "job_positions_edu = pd.DataFrame(mlb.transform(job_positions_train['Education']), columns=mlb.classes_)\n",
    "candidates_edu = pd.DataFrame(mlb.transform(candidates_train['Education']), columns=mlb.classes_)\n",
    "\n",
    "job_positions_edu_t = pd.DataFrame(mlb.transform(job_positions_test['Education']), columns=mlb.classes_)\n",
    "candidates_edu_t = pd.DataFrame(mlb.transform(candidates_test['Education']), columns=mlb.classes_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we created another dataset by dropping the old features and we substitutes them with the encoded one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Merge the encoded features back into the original DataFrames\n",
    "job_positions_train_encoded = pd.concat([job_positions_train.drop(['Hard_skills_group', 'Soft_skills', 'Languages', 'Cards','Education'], axis=1), \n",
    "                                         job_positions_hard_skills, job_positions_soft_skills, job_positions_languages, job_positions_cards,job_positions_edu], axis=1)\n",
    "\n",
    "candidates_train_encoded = pd.concat([candidates_train.drop(['Hard_skills_group', 'Soft_skills', 'Languages', 'Cards','Education'], axis=1), \n",
    "                                      candidates_hard_skills, candidates_soft_skills, candidates_languages, candidates_cards,candidates_edu], axis=1)\n",
    "\n",
    "# Merge the encoded features back into the original DataFrames\n",
    "job_positions_test_encoded = pd.concat([job_positions_test.drop(['Hard_skills_group', 'Soft_skills', 'Languages', 'Cards','Education'], axis=1), \n",
    "                                         job_positions_hard_skills_t, job_positions_soft_skills_t, job_positions_languages_t, job_positions_cards_t, job_positions_edu_t], axis=1)\n",
    "\n",
    "candidates_test_encoded = pd.concat([candidates_test.drop(['Hard_skills_group', 'Soft_skills', 'Languages', 'Cards','Education'], axis=1), \n",
    "                                      candidates_hard_skills_t, candidates_soft_skills_t, candidates_languages_t, candidates_cards_t, candidates_edu_t], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we needed to create another dataset, since after the encoding we have a lot of features. To create such dataset we decided to compute similaritites among features, in particular we took advantage of Jaccard and Cosine similarity."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to calculate Jaccard similarity\n",
    "def jaccard_similarity(set1, set2):\n",
    "    intersection = len(set1.intersection(set2))\n",
    "    union = len(set1.union(set2))\n",
    "    if union == 0:\n",
    "        return 0\n",
    "    return intersection / union\n",
    "\n",
    "# Calculate similarity features for a row\n",
    "def calculate_similarity_features(row, candidates_df, jobs_df):\n",
    "    candidate_id = row['CANDIDATE_ID']\n",
    "    job_id = row['Job_ID']\n",
    "    \n",
    "    candidate = candidates_df[candidates_df['CANDIDATE_ID'] == candidate_id].iloc[0]\n",
    "    job = jobs_df[jobs_df['Job_ID'] == job_id].iloc[0]\n",
    "    \n",
    "    features = {}\n",
    "    features['prof_cat_exp_similarity'] = jaccard_similarity(candidate['Prof_cat_exp'], set([job['Prof_cat']]))\n",
    "    features['prof_cat_int_similarity'] = jaccard_similarity(candidate['Prof_cat_int'], set([job['Prof_cat']]))\n",
    "    features['hard_skills_similarity'] = cosine_similarity([candidate[hard_skills_columns].values], [job[hard_skills_columns].values])[0][0]\n",
    "    features['soft_skills_similarity'] = cosine_similarity([candidate[soft_skills_columns].values], [job[soft_skills_columns].values])[0][0]\n",
    "    features['languages_similarity'] = cosine_similarity([candidate[languages_columns].values], [job[languages_columns].values])[0][0]\n",
    "    features['cards_similarity'] = cosine_similarity([candidate[cards_columns].values], [job[cards_columns].values])[0][0]\n",
    "    features['Education_similarity'] = cosine_similarity([candidate[education_columns].values], [job[education_columns].values])[0][0]\n",
    "    \n",
    "    # Combine feature vectors for overall similarity\n",
    "    candidate_vector = candidate[hard_skills_columns].tolist() + \\\n",
    "                       candidate[soft_skills_columns].tolist() + \\\n",
    "                       candidate[languages_columns].tolist() + \\\n",
    "                       candidate[cards_columns].tolist()\n",
    "    \n",
    "    job_vector = job[hard_skills_columns].tolist() + \\\n",
    "                 job[soft_skills_columns].tolist() + \\\n",
    "                 job[languages_columns].tolist() + \\\n",
    "                 job[cards_columns].tolist()\n",
    "    features['overall_similarity'] = cosine_similarity([candidate_vector], [job_vector])[0][0]\n",
    "\n",
    "    \n",
    "    return pd.Series(features)\n",
    "\n",
    "# Columns to calculate cosine similarity\n",
    "hard_skills_columns = candidates_hard_skills.columns\n",
    "soft_skills_columns = candidates_soft_skills.columns\n",
    "languages_columns = candidates_languages.columns\n",
    "cards_columns = candidates_cards.columns\n",
    "education_columns=candidates_edu.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that all our preprocessing is done, we need to create the final dataset. We have been given of a third dataset, called 'matches', which contains a job id, a candidate id and the flag 0/1 whether that candidate has or hasn't gotten the job. \n",
    "\n",
    "Notice that for each job there are 5 candidates but just 1 of them can be selected, but the same candidate may be selected for more than one job."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "matches_train = pd.read_csv('matches_train.csv')\n",
    "matches_test = pd.read_csv('matches_test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is how the matches dataset looks like:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>Match</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1933502</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0553_22671</td>\n",
       "      <td>6016478</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1354_730</td>\n",
       "      <td>6167354</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1526_995</td>\n",
       "      <td>5883294</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0130_11963</td>\n",
       "      <td>5848616</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12650</th>\n",
       "      <td>0178_21171</td>\n",
       "      <td>5011230</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12651</th>\n",
       "      <td>0070_20302</td>\n",
       "      <td>2214010</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12652</th>\n",
       "      <td>0349_15070</td>\n",
       "      <td>7638353</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12653</th>\n",
       "      <td>0509_16576</td>\n",
       "      <td>3451059</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12654</th>\n",
       "      <td>0327_12409</td>\n",
       "      <td>1002941</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12655 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Job_ID  CANDIDATE_ID  Match\n",
       "0      0264_22927       1933502      1\n",
       "1      0553_22671       6016478      1\n",
       "2        1354_730       6167354      1\n",
       "3        1526_995       5883294      1\n",
       "4      0130_11963       5848616      1\n",
       "...           ...           ...    ...\n",
       "12650  0178_21171       5011230      0\n",
       "12651  0070_20302       2214010      0\n",
       "12652  0349_15070       7638353      0\n",
       "12653  0509_16576       3451059      0\n",
       "12654  0327_12409       1002941      0\n",
       "\n",
       "[12655 rows x 3 columns]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our final dataset is based on the similarity among the features of the candidates and of the job position, which are then joined thanks to the 'matches' dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate similarity features\n",
    "similarity_features = matches_train.apply(lambda row: calculate_similarity_features(row, candidates_train_encoded, job_positions_train_encoded), axis=1)\n",
    "matches_train = matches_train.join(similarity_features)\n",
    "\n",
    "similarity_features = matches_test.apply(lambda row: calculate_similarity_features(row, candidates_test_encoded, job_positions_test_encoded), axis=1)\n",
    "matches_test = matches_test.join(similarity_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is how our new dataset looks like"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>Match</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1933502</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.258199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0553_22671</td>\n",
       "      <td>6016478</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1354_730</td>\n",
       "      <td>6167354</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.577350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1526_995</td>\n",
       "      <td>5883294</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0130_11963</td>\n",
       "      <td>5848616</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.176777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12650</th>\n",
       "      <td>0178_21171</td>\n",
       "      <td>5011230</td>\n",
       "      <td>0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.816497</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.239046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12651</th>\n",
       "      <td>0070_20302</td>\n",
       "      <td>2214010</td>\n",
       "      <td>0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.447214</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.400892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12652</th>\n",
       "      <td>0349_15070</td>\n",
       "      <td>7638353</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12653</th>\n",
       "      <td>0509_16576</td>\n",
       "      <td>3451059</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.577350</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.204124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12654</th>\n",
       "      <td>0327_12409</td>\n",
       "      <td>1002941</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12655 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Job_ID  CANDIDATE_ID  Match  prof_cat_exp_similarity  \\\n",
       "0      0264_22927       1933502      1                 1.000000   \n",
       "1      0553_22671       6016478      1                 1.000000   \n",
       "2        1354_730       6167354      1                 1.000000   \n",
       "3        1526_995       5883294      1                 1.000000   \n",
       "4      0130_11963       5848616      1                 1.000000   \n",
       "...           ...           ...    ...                      ...   \n",
       "12650  0178_21171       5011230      0                 0.500000   \n",
       "12651  0070_20302       2214010      0                 0.333333   \n",
       "12652  0349_15070       7638353      0                 0.000000   \n",
       "12653  0509_16576       3451059      0                 0.000000   \n",
       "12654  0327_12409       1002941      0                 0.000000   \n",
       "\n",
       "       prof_cat_int_similarity  hard_skills_similarity  \\\n",
       "0                          0.0                     0.0   \n",
       "1                          0.5                     0.0   \n",
       "2                          1.0                     1.0   \n",
       "3                          0.0                     0.0   \n",
       "4                          0.5                     0.0   \n",
       "...                        ...                     ...   \n",
       "12650                      0.0                     0.0   \n",
       "12651                      0.0                     0.5   \n",
       "12652                      0.0                     0.0   \n",
       "12653                      0.0                     0.0   \n",
       "12654                      0.0                     0.0   \n",
       "\n",
       "       soft_skills_similarity  languages_similarity  cards_similarity  \\\n",
       "0                    0.000000              0.000000               1.0   \n",
       "1                    0.000000              0.000000               0.0   \n",
       "2                    0.000000              0.000000               0.0   \n",
       "3                    0.000000              0.000000               0.0   \n",
       "4                    0.000000              0.000000               1.0   \n",
       "...                       ...                   ...               ...   \n",
       "12650                0.816497              0.000000               0.0   \n",
       "12651                0.447214              0.707107               0.0   \n",
       "12652                0.000000              0.000000               0.0   \n",
       "12653                0.577350              0.000000               0.0   \n",
       "12654                0.000000              0.000000               0.0   \n",
       "\n",
       "       Education_similarity  overall_similarity  \n",
       "0                       0.0            0.258199  \n",
       "1                       0.0            0.000000  \n",
       "2                       0.0            0.577350  \n",
       "3                       0.0            0.000000  \n",
       "4                       0.0            0.176777  \n",
       "...                     ...                 ...  \n",
       "12650                   0.0            0.239046  \n",
       "12651                   0.0            0.400892  \n",
       "12652                   0.0            0.000000  \n",
       "12653                   0.0            0.204124  \n",
       "12654                   0.0            0.000000  \n",
       "\n",
       "[12655 rows x 11 columns]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>Match</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1933502</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.258199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9274</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>2968728</td>\n",
       "      <td>0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.149071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9550</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>2643665</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.223607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10882</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1216748</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.169031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11113</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1441537</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.149071</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Job_ID  CANDIDATE_ID  Match  prof_cat_exp_similarity  \\\n",
       "0      0264_22927       1933502      1                 1.000000   \n",
       "9274   0264_22927       2968728      0                 0.333333   \n",
       "9550   0264_22927       2643665      0                 0.000000   \n",
       "10882  0264_22927       1216748      0                 0.000000   \n",
       "11113  0264_22927       1441537      0                 0.000000   \n",
       "\n",
       "       prof_cat_int_similarity  hard_skills_similarity  \\\n",
       "0                          0.0                     0.0   \n",
       "9274                       0.0                     0.0   \n",
       "9550                       0.0                     0.0   \n",
       "10882                      0.0                     0.0   \n",
       "11113                      0.0                     0.0   \n",
       "\n",
       "       soft_skills_similarity  languages_similarity  cards_similarity  \\\n",
       "0                         0.0                   0.0          1.000000   \n",
       "9274                      0.0                   0.0          0.707107   \n",
       "9550                      0.0                   0.0          1.000000   \n",
       "10882                     0.0                   0.0          1.000000   \n",
       "11113                     0.0                   0.0          0.707107   \n",
       "\n",
       "       Education_similarity  overall_similarity  \n",
       "0                       0.0            0.258199  \n",
       "9274                    0.0            0.149071  \n",
       "9550                    0.0            0.223607  \n",
       "10882                   0.0            0.169031  \n",
       "11113                   0.0            0.149071  "
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_train[matches_train.Job_ID == '0264_22927']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Match</th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>23262979</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.408248</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.316228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>1482016</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.182574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>83352824</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.316228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>54602123</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.408248</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.298142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>92641837</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.577350</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.134840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3995</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>82724002</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.365148</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.327327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3996</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>38957146</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3997</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>5137608</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.166667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3998</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>39393438</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.235702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3999</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>45683368</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4000 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Match    Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "0       NaN  1014_514      23262979                     1.00   \n",
       "1       NaN  1014_514       1482016                     0.25   \n",
       "2       NaN  1014_514      83352824                     0.50   \n",
       "3       NaN  1014_514      54602123                     1.00   \n",
       "4       NaN  1014_514      92641837                     0.25   \n",
       "...     ...       ...           ...                      ...   \n",
       "3995    NaN  9996_273      82724002                     0.00   \n",
       "3996    NaN  9996_273      38957146                     0.00   \n",
       "3997    NaN  9996_273       5137608                     0.00   \n",
       "3998    NaN  9996_273      39393438                     0.00   \n",
       "3999    NaN  9996_273      45683368                     0.50   \n",
       "\n",
       "      prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "0                    0.333333                0.408248                     0.0   \n",
       "1                    0.000000                0.000000                     0.0   \n",
       "2                    0.333333                1.000000                     0.0   \n",
       "3                    1.000000                0.408248                     0.0   \n",
       "4                    0.000000                0.000000                     0.0   \n",
       "...                       ...                     ...                     ...   \n",
       "3995                 0.000000                0.365148                     0.0   \n",
       "3996                 0.200000                0.000000                     0.0   \n",
       "3997                 0.000000                0.000000                     0.0   \n",
       "3998                 0.000000                0.000000                     0.0   \n",
       "3999                 0.333333                0.000000                     0.0   \n",
       "\n",
       "      languages_similarity  cards_similarity  Education_similarity  \\\n",
       "0                 1.000000          0.000000                   0.0   \n",
       "1                 0.707107          0.000000                   0.0   \n",
       "2                 0.000000          0.000000                   0.0   \n",
       "3                 1.000000          0.000000                   0.0   \n",
       "4                 0.577350          0.000000                   0.0   \n",
       "...                    ...               ...                   ...   \n",
       "3995              0.000000          0.707107                   0.0   \n",
       "3996              0.000000          0.000000                   0.0   \n",
       "3997              0.000000          1.000000                   0.0   \n",
       "3998              0.000000          0.707107                   0.0   \n",
       "3999              0.000000          1.000000                   0.0   \n",
       "\n",
       "      overall_similarity  \n",
       "0               0.316228  \n",
       "1               0.182574  \n",
       "2               0.316228  \n",
       "3               0.298142  \n",
       "4               0.134840  \n",
       "...                  ...  \n",
       "3995            0.327327  \n",
       "3996            0.000000  \n",
       "3997            0.166667  \n",
       "3998            0.235702  \n",
       "3999            0.408248  \n",
       "\n",
       "[4000 rows x 11 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also noticed that 'Latitudine' and 'Longitudine' have missing values, therefore we imputed them by scraping the coordinates of the city of the missing entries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize the Nominatim geocoder\n",
    "geolocator = Nominatim(user_agent=\"geoapiExercises\")\n",
    "\n",
    "# Function to get latitude and longitude\n",
    "def get_lat_lon(city_name):\n",
    "    location = geolocator.geocode(city_name)\n",
    "    if location:\n",
    "        return location.latitude, location.longitude\n",
    "    else:\n",
    "        return None, None"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Impute missing values:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Iterate through the DataFrame and update missing values\n",
    "for idx, row in candidates_train.iterrows():\n",
    "    if row['Latitudine'] == '' or row['Longitudine'] == '':\n",
    "        city = row['Domicile_City']\n",
    "        latitude, longitude = get_lat_lon(city)\n",
    "        candidates_train.at[idx, 'Latitudine'] = latitude\n",
    "        candidates_train.at[idx, 'Longitudine'] = longitude\n",
    "        # Pause for a moment to respect the usage policy of the geocoding service\n",
    "        time.sleep(1)\n",
    "\n",
    "for idx, row in job_positions_train.iterrows():\n",
    "    if row['Latitudine'] == '' or row['Longitudine'] == '':\n",
    "        city = row['Work_city']\n",
    "        latitude, longitude = get_lat_lon(city)\n",
    "        job_positions_train.at[idx, 'Latitudine'] = latitude\n",
    "        job_positions_train.at[idx, 'Longitudine'] = longitude\n",
    "        # Pause for a moment to respect the usage policy of the geocoding service\n",
    "        time.sleep(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Iterate through the DataFrame and update missing values\n",
    "for idx, row in candidates_test.iterrows():\n",
    "    if row['Latitudine'] == '' or row['Longitudine'] == '':\n",
    "        city = row['Domicile_City']\n",
    "        latitude, longitude = get_lat_lon(city)\n",
    "        candidates_test.at[idx, 'Latitudine'] = latitude\n",
    "        candidates_test.at[idx, 'Longitudine'] = longitude\n",
    "        # Pause for a moment to respect the usage policy of the geocoding service\n",
    "        time.sleep(1)\n",
    "\n",
    "for idx, row in job_positions_test.iterrows():\n",
    "    if row['Latitudine'] == '' or row['Longitudine'] == '':\n",
    "        city = row['Work_city']\n",
    "        latitude, longitude = get_lat_lon(city)\n",
    "        job_positions_test.at[idx, 'Latitudine'] = latitude\n",
    "        job_positions_test.at[idx, 'Longitudine'] = longitude\n",
    "        # Pause for a moment to respect the usage policy of the geocoding service\n",
    "        time.sleep(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add latitude and longitude to our dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "cand_to_merge = candidates_train[['CANDIDATE_ID', 'Latitudine', 'Longitudine']]\n",
    "job_to_merge = job_positions_train[['Job_ID', 'Latitudine', 'Longitudine']]\n",
    "\n",
    "cand_to_merge_test = candidates_test[['CANDIDATE_ID', 'Latitudine', 'Longitudine']]\n",
    "job_to_merge_test = job_positions_test[['Job_ID', 'Latitudine', 'Longitudine']]\n",
    "\n",
    "matches_train_d = matches_train.merge(cand_to_merge, on='CANDIDATE_ID', how='left').merge(job_to_merge, on='Job_ID', how='left')\n",
    "matches_test_d = matches_test.merge(cand_to_merge_test, on='CANDIDATE_ID', how='left').merge(job_to_merge_test, on='Job_ID', how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To compute the distance between the Candidate city and the Job city we took advantage of the following function that allowed us to compute the circle distance in kilometers between two points on the earth."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "def haversine(lat1, lon1, lat2, lon2):\n",
    "    \"\"\"\n",
    "    Calculate the great circle distance in kilometers between two points \n",
    "    on the earth (specified in decimal degrees)\n",
    "    \"\"\"\n",
    "    lat1 = float(lat1)\n",
    "    lon1 = float(lon1)\n",
    "    lat2 = float(lat2)\n",
    "    lon2 = float(lon2)\n",
    "    # Convert latitude and longitude from degrees to radians\n",
    "    lat1, lon1, lat2, lon2 = map(radians, [lat1, lon1, lat2, lon2])\n",
    "    \n",
    "    # Haversine formula\n",
    "    dlon = lon2 - lon1\n",
    "    dlat = lat2 - lat1\n",
    "    a = sin(dlat/2)**2 + cos(lat1) * cos(lat2) * sin(dlon/2)**2\n",
    "    c = 2 * atan2(sqrt(a), sqrt(1 - a))\n",
    "    \n",
    "    # Radius of the Earth in kilometers\n",
    "    R = 6371.0\n",
    "    \n",
    "    # Calculate the distance\n",
    "    distance = R * c\n",
    "    \n",
    "    return distance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compute such distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "matches_train['distance'] = matches_train_d.apply(lambda row: haversine(row['Latitudine_x'], row['Longitudine_x'], row['Latitudine_y'], row['Longitudine_y']), axis=1)\n",
    "matches_test['distance'] = matches_test_d.apply(lambda row: haversine(row['Latitudine_x'], row['Longitudine_x'], row['Latitudine_y'], row['Longitudine_y']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>Match</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1933502</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.258199</td>\n",
       "      <td>2.132456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0553_22671</td>\n",
       "      <td>6016478</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.767585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1354_730</td>\n",
       "      <td>6167354</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.577350</td>\n",
       "      <td>6.301685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1526_995</td>\n",
       "      <td>5883294</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>16.536885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0130_11963</td>\n",
       "      <td>5848616</td>\n",
       "      <td>1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.176777</td>\n",
       "      <td>6.732852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12650</th>\n",
       "      <td>0178_21171</td>\n",
       "      <td>5011230</td>\n",
       "      <td>0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.816497</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.239046</td>\n",
       "      <td>29.687561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12651</th>\n",
       "      <td>0070_20302</td>\n",
       "      <td>2214010</td>\n",
       "      <td>0</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.447214</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.400892</td>\n",
       "      <td>25.380609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12652</th>\n",
       "      <td>0349_15070</td>\n",
       "      <td>7638353</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.991345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12653</th>\n",
       "      <td>0509_16576</td>\n",
       "      <td>3451059</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.577350</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.204124</td>\n",
       "      <td>12.046100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12654</th>\n",
       "      <td>0327_12409</td>\n",
       "      <td>1002941</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>42.362678</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12655 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Job_ID  CANDIDATE_ID  Match  prof_cat_exp_similarity  \\\n",
       "0      0264_22927       1933502      1                 1.000000   \n",
       "1      0553_22671       6016478      1                 1.000000   \n",
       "2        1354_730       6167354      1                 1.000000   \n",
       "3        1526_995       5883294      1                 1.000000   \n",
       "4      0130_11963       5848616      1                 1.000000   \n",
       "...           ...           ...    ...                      ...   \n",
       "12650  0178_21171       5011230      0                 0.500000   \n",
       "12651  0070_20302       2214010      0                 0.333333   \n",
       "12652  0349_15070       7638353      0                 0.000000   \n",
       "12653  0509_16576       3451059      0                 0.000000   \n",
       "12654  0327_12409       1002941      0                 0.000000   \n",
       "\n",
       "       prof_cat_int_similarity  hard_skills_similarity  \\\n",
       "0                          0.0                     0.0   \n",
       "1                          0.5                     0.0   \n",
       "2                          1.0                     1.0   \n",
       "3                          0.0                     0.0   \n",
       "4                          0.5                     0.0   \n",
       "...                        ...                     ...   \n",
       "12650                      0.0                     0.0   \n",
       "12651                      0.0                     0.5   \n",
       "12652                      0.0                     0.0   \n",
       "12653                      0.0                     0.0   \n",
       "12654                      0.0                     0.0   \n",
       "\n",
       "       soft_skills_similarity  languages_similarity  cards_similarity  \\\n",
       "0                    0.000000              0.000000               1.0   \n",
       "1                    0.000000              0.000000               0.0   \n",
       "2                    0.000000              0.000000               0.0   \n",
       "3                    0.000000              0.000000               0.0   \n",
       "4                    0.000000              0.000000               1.0   \n",
       "...                       ...                   ...               ...   \n",
       "12650                0.816497              0.000000               0.0   \n",
       "12651                0.447214              0.707107               0.0   \n",
       "12652                0.000000              0.000000               0.0   \n",
       "12653                0.577350              0.000000               0.0   \n",
       "12654                0.000000              0.000000               0.0   \n",
       "\n",
       "       Education_similarity  overall_similarity   distance  \n",
       "0                       0.0            0.258199   2.132456  \n",
       "1                       0.0            0.000000   8.767585  \n",
       "2                       0.0            0.577350   6.301685  \n",
       "3                       0.0            0.000000  16.536885  \n",
       "4                       0.0            0.176777   6.732852  \n",
       "...                     ...                 ...        ...  \n",
       "12650                   0.0            0.239046  29.687561  \n",
       "12651                   0.0            0.400892  25.380609  \n",
       "12652                   0.0            0.000000   9.991345  \n",
       "12653                   0.0            0.204124  12.046100  \n",
       "12654                   0.0            0.000000  42.362678  \n",
       "\n",
       "[12655 rows x 12 columns]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add also the squared distance, to have more features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "matches_train['distance2'] = matches_train['distance']**2\n",
    "matches_test['distance2'] = matches_test['distance']**2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Of course we need to scale our data to get a better representation and improve our future models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize StandardScaler\n",
    "scaler = StandardScaler()\n",
    "\n",
    "features_to_scale = ['prof_cat_exp_similarity', 'prof_cat_int_similarity', \n",
    "                     'hard_skills_similarity', 'soft_skills_similarity', 'languages_similarity', 'cards_similarity',\n",
    "                     'distance', 'Education_similarity', 'overall_similarity', 'distance2']\n",
    "\n",
    "# Scale the dataset\n",
    "scaled_data = scaler.fit_transform(matches_train[features_to_scale])\n",
    "\n",
    "scaled_data_test = scaler.fit_transform(matches_test[features_to_scale])\n",
    "\n",
    "# Convert to DataFrame for readability\n",
    "scaled_df = pd.DataFrame(scaled_data, columns=matches_train[features_to_scale].columns)\n",
    "scaled_df_test = pd.DataFrame(scaled_data_test, columns=matches_test[features_to_scale].columns)\n",
    "\n",
    "matches_train_scaled = pd.concat([matches_train.drop(features_to_scale, axis=1), scaled_df], axis=1)\n",
    "matches_test_scaled = pd.concat([matches_test.drop(features_to_scale, axis=1), scaled_df_test], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quick look to how the scaled dataset looks like"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>Match</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0264_22927</td>\n",
       "      <td>1933502</td>\n",
       "      <td>1</td>\n",
       "      <td>1.792094</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>2.177872</td>\n",
       "      <td>-0.226177</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>0.289542</td>\n",
       "      <td>-0.035418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0553_22671</td>\n",
       "      <td>6016478</td>\n",
       "      <td>1</td>\n",
       "      <td>1.792094</td>\n",
       "      <td>0.841365</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.201563</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>-0.905886</td>\n",
       "      <td>-0.035385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1354_730</td>\n",
       "      <td>6167354</td>\n",
       "      <td>1</td>\n",
       "      <td>1.792094</td>\n",
       "      <td>2.275002</td>\n",
       "      <td>2.857453</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.210711</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>1.767172</td>\n",
       "      <td>-0.035402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1526_995</td>\n",
       "      <td>5883294</td>\n",
       "      <td>1</td>\n",
       "      <td>1.792094</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.172742</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>-0.905886</td>\n",
       "      <td>-0.035294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0130_11963</td>\n",
       "      <td>5848616</td>\n",
       "      <td>1</td>\n",
       "      <td>1.792094</td>\n",
       "      <td>0.841365</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>2.177872</td>\n",
       "      <td>-0.209111</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>-0.087432</td>\n",
       "      <td>-0.035400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12650</th>\n",
       "      <td>0178_21171</td>\n",
       "      <td>5011230</td>\n",
       "      <td>0</td>\n",
       "      <td>0.558762</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>2.756856</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.123959</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>0.200866</td>\n",
       "      <td>-0.035013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12651</th>\n",
       "      <td>0070_20302</td>\n",
       "      <td>2214010</td>\n",
       "      <td>0</td>\n",
       "      <td>0.147652</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>1.142643</td>\n",
       "      <td>1.276256</td>\n",
       "      <td>1.907498</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.139936</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>0.950192</td>\n",
       "      <td>-0.035123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12652</th>\n",
       "      <td>0349_15070</td>\n",
       "      <td>7638353</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.674570</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.197023</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>-0.905886</td>\n",
       "      <td>-0.035374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12653</th>\n",
       "      <td>0509_16576</td>\n",
       "      <td>3451059</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.674570</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>1.798025</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.189401</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>0.039183</td>\n",
       "      <td>-0.035353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12654</th>\n",
       "      <td>0327_12409</td>\n",
       "      <td>1002941</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.674570</td>\n",
       "      <td>-0.592272</td>\n",
       "      <td>-0.572167</td>\n",
       "      <td>-0.516798</td>\n",
       "      <td>-0.437074</td>\n",
       "      <td>-0.530162</td>\n",
       "      <td>-0.076939</td>\n",
       "      <td>-0.102312</td>\n",
       "      <td>-0.905886</td>\n",
       "      <td>-0.034591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12655 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Job_ID  CANDIDATE_ID  Match  prof_cat_exp_similarity  \\\n",
       "0      0264_22927       1933502      1                 1.792094   \n",
       "1      0553_22671       6016478      1                 1.792094   \n",
       "2        1354_730       6167354      1                 1.792094   \n",
       "3        1526_995       5883294      1                 1.792094   \n",
       "4      0130_11963       5848616      1                 1.792094   \n",
       "...           ...           ...    ...                      ...   \n",
       "12650  0178_21171       5011230      0                 0.558762   \n",
       "12651  0070_20302       2214010      0                 0.147652   \n",
       "12652  0349_15070       7638353      0                -0.674570   \n",
       "12653  0509_16576       3451059      0                -0.674570   \n",
       "12654  0327_12409       1002941      0                -0.674570   \n",
       "\n",
       "       prof_cat_int_similarity  hard_skills_similarity  \\\n",
       "0                    -0.592272               -0.572167   \n",
       "1                     0.841365               -0.572167   \n",
       "2                     2.275002                2.857453   \n",
       "3                    -0.592272               -0.572167   \n",
       "4                     0.841365               -0.572167   \n",
       "...                        ...                     ...   \n",
       "12650                -0.592272               -0.572167   \n",
       "12651                -0.592272                1.142643   \n",
       "12652                -0.592272               -0.572167   \n",
       "12653                -0.592272               -0.572167   \n",
       "12654                -0.592272               -0.572167   \n",
       "\n",
       "       soft_skills_similarity  languages_similarity  cards_similarity  \\\n",
       "0                   -0.516798             -0.437074          2.177872   \n",
       "1                   -0.516798             -0.437074         -0.530162   \n",
       "2                   -0.516798             -0.437074         -0.530162   \n",
       "3                   -0.516798             -0.437074         -0.530162   \n",
       "4                   -0.516798             -0.437074          2.177872   \n",
       "...                       ...                   ...               ...   \n",
       "12650                2.756856             -0.437074         -0.530162   \n",
       "12651                1.276256              1.907498         -0.530162   \n",
       "12652               -0.516798             -0.437074         -0.530162   \n",
       "12653                1.798025             -0.437074         -0.530162   \n",
       "12654               -0.516798             -0.437074         -0.530162   \n",
       "\n",
       "       distance  Education_similarity  overall_similarity  distance2  \n",
       "0     -0.226177             -0.102312            0.289542  -0.035418  \n",
       "1     -0.201563             -0.102312           -0.905886  -0.035385  \n",
       "2     -0.210711             -0.102312            1.767172  -0.035402  \n",
       "3     -0.172742             -0.102312           -0.905886  -0.035294  \n",
       "4     -0.209111             -0.102312           -0.087432  -0.035400  \n",
       "...         ...                   ...                 ...        ...  \n",
       "12650 -0.123959             -0.102312            0.200866  -0.035013  \n",
       "12651 -0.139936             -0.102312            0.950192  -0.035123  \n",
       "12652 -0.197023             -0.102312           -0.905886  -0.035374  \n",
       "12653 -0.189401             -0.102312            0.039183  -0.035353  \n",
       "12654 -0.076939             -0.102312           -0.905886  -0.034591  \n",
       "\n",
       "[12655 rows x 13 columns]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_train_scaled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Match</th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>23262979</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>0.887062</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.092271</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>1482016</td>\n",
       "      <td>-0.062095</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>1.832306</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.092271</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.045517</td>\n",
       "      <td>-0.045888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>83352824</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.054167</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>54602123</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>2.226710</td>\n",
       "      <td>0.887062</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.055541</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.497398</td>\n",
       "      <td>-0.045512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1014_514</td>\n",
       "      <td>92641837</td>\n",
       "      <td>-0.062095</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.033892</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.269762</td>\n",
       "      <td>-0.045233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3995</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>82724002</td>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>0.733830</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.117389</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.634500</td>\n",
       "      <td>-0.046077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3996</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>38957146</td>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.032375</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.085699</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.045830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3997</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>5137608</td>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.167453</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.120247</td>\n",
       "      <td>-0.046283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3998</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>39393438</td>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.136425</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.204067</td>\n",
       "      <td>-0.046181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3999</th>\n",
       "      <td>NaN</td>\n",
       "      <td>9996_273</td>\n",
       "      <td>45683368</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.182602</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.014652</td>\n",
       "      <td>-0.046301</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4000 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Match    Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "0       NaN  1014_514      23262979                 1.789496   \n",
       "1       NaN  1014_514       1482016                -0.062095   \n",
       "2       NaN  1014_514      83352824                 0.555102   \n",
       "3       NaN  1014_514      54602123                 1.789496   \n",
       "4       NaN  1014_514      92641837                -0.062095   \n",
       "...     ...       ...           ...                      ...   \n",
       "3995    NaN  9996_273      82724002                -0.679291   \n",
       "3996    NaN  9996_273      38957146                -0.679291   \n",
       "3997    NaN  9996_273       5137608                -0.679291   \n",
       "3998    NaN  9996_273      39393438                -0.679291   \n",
       "3999    NaN  9996_273      45683368                 0.555102   \n",
       "\n",
       "      prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "0                    0.344139                0.887062                -0.51756   \n",
       "1                   -0.597146               -0.564377                -0.51756   \n",
       "2                    0.344139                2.990909                -0.51756   \n",
       "3                    2.226710                0.887062                -0.51756   \n",
       "4                   -0.597146               -0.564377                -0.51756   \n",
       "...                       ...                     ...                     ...   \n",
       "3995                -0.597146                0.733830                -0.51756   \n",
       "3996                -0.032375               -0.564377                -0.51756   \n",
       "3997                -0.597146               -0.564377                -0.51756   \n",
       "3998                -0.597146               -0.564377                -0.51756   \n",
       "3999                 0.344139               -0.564377                -0.51756   \n",
       "\n",
       "      languages_similarity  cards_similarity  distance  Education_similarity  \\\n",
       "0                 2.776838         -0.519310 -0.092271             -0.022366   \n",
       "1                 1.832306         -0.519310 -0.092271             -0.022366   \n",
       "2                -0.447997         -0.519310 -0.054167             -0.022366   \n",
       "3                 2.776838         -0.519310 -0.055541             -0.022366   \n",
       "4                 1.413862         -0.519310 -0.033892             -0.022366   \n",
       "...                    ...               ...       ...                   ...   \n",
       "3995             -0.447997          1.426400 -0.117389             -0.022366   \n",
       "3996             -0.447997         -0.519310 -0.085699             -0.022366   \n",
       "3997             -0.447997          2.232339 -0.167453             -0.022366   \n",
       "3998             -0.447997          1.426400 -0.136425             -0.022366   \n",
       "3999             -0.447997          2.232339 -0.182602             -0.022366   \n",
       "\n",
       "      overall_similarity  distance2  \n",
       "0               0.582359  -0.045888  \n",
       "1              -0.045517  -0.045888  \n",
       "2               0.582359  -0.045495  \n",
       "3               0.497398  -0.045512  \n",
       "4              -0.269762  -0.045233  \n",
       "...                  ...        ...  \n",
       "3995            0.634500  -0.046077  \n",
       "3996           -0.903212  -0.045830  \n",
       "3997           -0.120247  -0.046283  \n",
       "3998            0.204067  -0.046181  \n",
       "3999            1.014652  -0.046301  \n",
       "\n",
       "[4000 rows x 13 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matches_test_scaled"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the indexes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "matches_train_scaled.set_index(['Job_ID'], inplace=True)\n",
    "matches_test_scaled.set_index(['Job_ID', 'CANDIDATE_ID'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#matches_train_scaled.to_csv('matches_train_scaled1.csv')\n",
    "#matches_test_scaled.to_csv('matches_test_scaled1.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After the long preprocessing, we finally obtained our final dataset, therefore we could start working on it. Our first naive approach was a simple LogisticRegression."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = matches_train_scaled.drop(['Match', 'CANDIDATE_ID'], axis=1)\n",
    "y = matches_train_scaled['Match']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy: 0.4970367443698143\n",
      "Placement Accuracy: 0.9163424124513618\n"
     ]
    }
   ],
   "source": [
    "# Split the data into training and testing sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=13)\n",
    "\n",
    "# Define the parameter grid for GridSearchCV\n",
    "param_grid = {\n",
    "    'C': [0.1, 1, 10, 100],  # Regularization parameter\n",
    "    'solver': ['liblinear', 'lbfgs'],  # Algorithm to use in the optimization problem\n",
    "    'max_iter': [100, 200, 300]  # Maximum number of iterations for optimization\n",
    "}\n",
    "\n",
    "# Initialize Logistic Regression model\n",
    "model = LogisticRegression()\n",
    "\n",
    "# Initialize GridSearchCV with the model and parameter grid\n",
    "grid_search = GridSearchCV(estimator=model, param_grid=param_grid, cv=5, scoring='accuracy')\n",
    "\n",
    "# Fit GridSearchCV on training data to find the best parameters\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "# Get the best model from GridSearchCV\n",
    "best_model = grid_search.best_estimator_\n",
    "\n",
    "# Predict probabilities on the test set using the best model\n",
    "y_pred = best_model.predict_proba(X_test)[:, 1]\n",
    "\n",
    "# Create a DataFrame with predictions and actual labels\n",
    "X_test_pred = X_test.copy()\n",
    "X_test_pred['predicted'] = y_pred\n",
    "X_test_pred['actual'] = y_test\n",
    "\n",
    "# Apply the 'Max' rule based on predictions\n",
    "X_test_pred['Max'] = X_test_pred.groupby('Job_ID')['predicted'].transform(lambda x: (x == x.max()).astype(int))\n",
    "\n",
    "# Filter rows where actual label is 1 (placement)\n",
    "X_tt = X_test_pred[X_test_pred['actual'] == 1]\n",
    "\n",
    "# Evaluate accuracy scores\n",
    "overall_accuracy = accuracy_score(X_test_pred['actual'], X_test_pred['Max'])\n",
    "placement_accuracy = accuracy_score(X_tt['actual'], X_tt['Max'])\n",
    "\n",
    "print(f\"Overall Accuracy: {overall_accuracy}\")\n",
    "print(f\"Placement Accuracy: {placement_accuracy}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we tried a gradient boos model, the XGBClassifier:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 108 candidates, totalling 540 fits\n",
      "0.4037929672066377\n",
      "0.9260700389105059\n"
     ]
    }
   ],
   "source": [
    "# Initialize the XGBoost classifier\n",
    "xgb_classifier = XGBClassifier(use_label_encoder=False, eval_metric='mlogloss')\n",
    "\n",
    "param_grid = {\n",
    "    'n_estimators': [50, 100, 150],\n",
    "    'max_depth': [3, 4, 5],\n",
    "    'learning_rate': [0.01, 0.1, 0.2],\n",
    "    'subsample': [0.8, 1.0],\n",
    "    'colsample_bytree': [0.8, 1.0]\n",
    "}\n",
    "# Train the model\n",
    "grid_search = GridSearchCV(estimator=xgb_classifier, param_grid=param_grid, cv=5, scoring='accuracy', n_jobs=-1, verbose=2)\n",
    "grid_search.fit(X_train, y_train)\n",
    "\n",
    "# Get the best model\n",
    "best_model = grid_search.best_estimator_\n",
    "\n",
    "# Train the best model on the full training data\n",
    "best_model.fit(X_train, y_train)\n",
    "# Make predictions on the test set\n",
    "y_pred = best_model.predict_proba(X_test)[:, 1]\n",
    "\n",
    "X_test_pred = X_test.copy()\n",
    "X_test_pred['predicted'] = y_pred\n",
    "X_test_pred['actual'] = y_test\n",
    "\n",
    "X_test_pred['Max'] = X_test_pred.groupby('Job_ID')['predicted'].transform(lambda x: (x == x.max()).astype(int))\n",
    "X_tt = X_test_pred[X_test_pred['actual'] == 1]\n",
    "\n",
    "print(accuracy_score(X_test_pred['actual'], X_test_pred['Max']))\n",
    "print(accuracy_score(X_tt['actual'], X_tt['Max']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After some tests, we landed on the Neural Network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_6 (Dense)             (None, 64)                704       \n",
      "                                                                 \n",
      " dropout_4 (Dropout)         (None, 64)                0         \n",
      "                                                                 \n",
      " dense_7 (Dense)             (None, 32)                2080      \n",
      "                                                                 \n",
      " dropout_5 (Dropout)         (None, 32)                0         \n",
      "                                                                 \n",
      " dense_8 (Dense)             (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2817 (11.00 KB)\n",
      "Trainable params: 2817 (11.00 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/20\n",
      "317/317 [==============================] - 3s 5ms/step - loss: 0.3953 - accuracy: 0.8024 - val_loss: 0.2869 - val_accuracy: 0.8352\n",
      "Epoch 2/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2841 - accuracy: 0.8486 - val_loss: 0.2665 - val_accuracy: 0.8412\n",
      "Epoch 3/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2660 - accuracy: 0.8452 - val_loss: 0.2614 - val_accuracy: 0.8431\n",
      "Epoch 4/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2590 - accuracy: 0.8485 - val_loss: 0.2548 - val_accuracy: 0.8455\n",
      "Epoch 5/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2519 - accuracy: 0.8522 - val_loss: 0.2536 - val_accuracy: 0.8459\n",
      "Epoch 6/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2492 - accuracy: 0.8490 - val_loss: 0.2528 - val_accuracy: 0.8459\n",
      "Epoch 7/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2482 - accuracy: 0.8494 - val_loss: 0.2525 - val_accuracy: 0.8435\n",
      "Epoch 8/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2463 - accuracy: 0.8510 - val_loss: 0.2521 - val_accuracy: 0.8463\n",
      "Epoch 9/20\n",
      "317/317 [==============================] - 1s 4ms/step - loss: 0.2457 - accuracy: 0.8540 - val_loss: 0.2525 - val_accuracy: 0.8424\n",
      "Epoch 10/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2440 - accuracy: 0.8498 - val_loss: 0.2523 - val_accuracy: 0.8439\n",
      "Epoch 11/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2431 - accuracy: 0.8503 - val_loss: 0.2507 - val_accuracy: 0.8447\n",
      "Epoch 12/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2441 - accuracy: 0.8506 - val_loss: 0.2503 - val_accuracy: 0.8459\n",
      "Epoch 13/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2417 - accuracy: 0.8509 - val_loss: 0.2495 - val_accuracy: 0.8451\n",
      "Epoch 14/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2394 - accuracy: 0.8540 - val_loss: 0.2499 - val_accuracy: 0.8455\n",
      "Epoch 15/20\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2406 - accuracy: 0.8557 - val_loss: 0.2494 - val_accuracy: 0.8443\n",
      "Epoch 16/20\n",
      "317/317 [==============================] - 2s 5ms/step - loss: 0.2409 - accuracy: 0.8548 - val_loss: 0.2502 - val_accuracy: 0.8424\n",
      "Epoch 17/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2411 - accuracy: 0.8519 - val_loss: 0.2507 - val_accuracy: 0.8431\n",
      "Epoch 18/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2400 - accuracy: 0.8570 - val_loss: 0.2498 - val_accuracy: 0.8451\n",
      "Epoch 19/20\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2392 - accuracy: 0.8537 - val_loss: 0.2500 - val_accuracy: 0.8443\n",
      "Epoch 20/20\n",
      "317/317 [==============================] - 1s 4ms/step - loss: 0.2386 - accuracy: 0.8559 - val_loss: 0.2493 - val_accuracy: 0.8439\n",
      "80/80 [==============================] - 0s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "# Define the neural network architecture\n",
    "model = Sequential([\n",
    "    Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    Dropout(0.5),\n",
    "    Dense(32, activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(1, activation='sigmoid')\n",
    "])\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer=Adam(learning_rate=0.001),\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# Print a summary of the model architecture\n",
    "model.summary()\n",
    "\n",
    "# Train the model\n",
    "history = model.fit(X_train, y_train,\n",
    "                    epochs=20,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test, y_test),\n",
    "                    verbose=1)\n",
    "\n",
    "# Predict probabilities on the test set\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_19\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_57 (Dense)            (None, 128)               1408      \n",
      "                                                                 \n",
      " dropout_38 (Dropout)        (None, 128)               0         \n",
      "                                                                 \n",
      " dense_58 (Dense)            (None, 64)                8256      \n",
      "                                                                 \n",
      " dropout_39 (Dropout)        (None, 64)                0         \n",
      "                                                                 \n",
      " dense_59 (Dense)            (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 9729 (38.00 KB)\n",
      "Trainable params: 9729 (38.00 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n",
      "Epoch 1/50\n",
      "317/317 [==============================] - 2s 3ms/step - loss: 0.3355 - accuracy: 0.8307 - val_loss: 0.2754 - val_accuracy: 0.8392\n",
      "Epoch 2/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2634 - accuracy: 0.8474 - val_loss: 0.2603 - val_accuracy: 0.8443\n",
      "Epoch 3/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2518 - accuracy: 0.8504 - val_loss: 0.2550 - val_accuracy: 0.8463\n",
      "Epoch 4/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2483 - accuracy: 0.8512 - val_loss: 0.2546 - val_accuracy: 0.8435\n",
      "Epoch 5/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2430 - accuracy: 0.8505 - val_loss: 0.2523 - val_accuracy: 0.8447\n",
      "Epoch 6/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2439 - accuracy: 0.8533 - val_loss: 0.2528 - val_accuracy: 0.8483\n",
      "Epoch 7/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2423 - accuracy: 0.8532 - val_loss: 0.2531 - val_accuracy: 0.8459\n",
      "Epoch 8/50\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2399 - accuracy: 0.8534 - val_loss: 0.2495 - val_accuracy: 0.8451\n",
      "Epoch 9/50\n",
      "317/317 [==============================] - 1s 4ms/step - loss: 0.2408 - accuracy: 0.8521 - val_loss: 0.2501 - val_accuracy: 0.8416\n",
      "Epoch 10/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2394 - accuracy: 0.8559 - val_loss: 0.2496 - val_accuracy: 0.8455\n",
      "Epoch 11/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2387 - accuracy: 0.8572 - val_loss: 0.2493 - val_accuracy: 0.8495\n",
      "Epoch 12/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2390 - accuracy: 0.8582 - val_loss: 0.2507 - val_accuracy: 0.8467\n",
      "Epoch 13/50\n",
      "317/317 [==============================] - 1s 2ms/step - loss: 0.2379 - accuracy: 0.8561 - val_loss: 0.2503 - val_accuracy: 0.8431\n",
      "Epoch 14/50\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2384 - accuracy: 0.8567 - val_loss: 0.2493 - val_accuracy: 0.8503\n",
      "Epoch 15/50\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2359 - accuracy: 0.8604 - val_loss: 0.2507 - val_accuracy: 0.8475\n",
      "Epoch 16/50\n",
      "317/317 [==============================] - 1s 3ms/step - loss: 0.2376 - accuracy: 0.8561 - val_loss: 0.2509 - val_accuracy: 0.8491\n",
      "80/80 [==============================] - 0s 1ms/step\n"
     ]
    }
   ],
   "source": [
    "# Split the data into training and testing sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=113)\n",
    "\n",
    "# Define the neural network architecture\n",
    "model = Sequential([\n",
    "    Dense(128, activation='relu', input_shape=(X_train.shape[1],)),\n",
    "    Dropout(0.5),\n",
    "    Dense(64, activation='relu',),\n",
    "    Dropout(0.5),\n",
    "    Dense(1, activation='sigmoid')\n",
    "])\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer=Adam(learning_rate=0.001),\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# Print a summary of the model architecture\n",
    "model.summary()\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=5, restore_best_weights=True)\n",
    "\n",
    "history = model.fit(X_train, y_train,\n",
    "                    epochs=50,  # Increase epochs for more training if needed\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test, y_test),\n",
    "                    callbacks=[early_stopping],\n",
    "                    verbose=1)\n",
    "\n",
    "# Predict probabilities on the test set\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting the training and validation accuracy and loss\n",
    "plt.figure(figsize=(12, 6))\n",
    "plt.rcParams.update({\n",
    "    'font.size': 14,\n",
    "    'axes.facecolor': '#f2f2f2',  # Very light grey background\n",
    "    'grid.color': 'white',\n",
    "    'axes.edgecolor': '#f8f8f8',  # Removing border\n",
    "    'axes.linewidth': 0.0\n",
    "})\n",
    "\n",
    "# Plot accuracy\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(history.history['accuracy'], color='red', label='Train Accuracy')\n",
    "plt.plot(history.history['val_accuracy'], color='purple', label='Validation Accuracy')\n",
    "plt.title('Model Accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(loc='best')\n",
    "plt.grid(True)\n",
    "\n",
    "# Plot loss\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(history.history['loss'], color='red', label='Train Loss')\n",
    "plt.plot(history.history['val_loss'], color='purple', label='Validation Loss')\n",
    "plt.title('Model Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend(loc='best')\n",
    "plt.grid(True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.49861714737258\n",
      "0.9202334630350194\n"
     ]
    }
   ],
   "source": [
    "X_test_pred = X_test.copy()\n",
    "X_test_pred['predicted'] = y_pred\n",
    "X_test_pred['actual'] = y_test\n",
    "\n",
    "X_test_pred['Max'] = X_test_pred.groupby('Job_ID')['predicted'].transform(lambda x: (x == x.max()).astype(int))\n",
    "X_tt = X_test_pred[X_test_pred['actual'] == 1]\n",
    "\n",
    "print(accuracy_score(X_test_pred['actual'], X_test_pred['Max']))\n",
    "print(accuracy_score(X_tt['actual'], X_tt['Max']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = matches_test_scaled.drop(['Match'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  1/125 [..............................] - ETA: 3s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "125/125 [==============================] - 0s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "# Predict probabilities on the test set\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test['prediction'] = y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "800"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.index.get_level_values(0).nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">1014_514</th>\n",
       "      <th>23262979</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>0.887062</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.092271</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045888</td>\n",
       "      <td>0.618107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1482016</th>\n",
       "      <td>-0.062095</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>1.832306</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.092271</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.045517</td>\n",
       "      <td>-0.045888</td>\n",
       "      <td>0.436729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83352824</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.054167</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045495</td>\n",
       "      <td>0.638060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54602123</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>2.226710</td>\n",
       "      <td>0.887062</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.055541</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.497398</td>\n",
       "      <td>-0.045512</td>\n",
       "      <td>0.551775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92641837</th>\n",
       "      <td>-0.062095</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.033892</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.269762</td>\n",
       "      <td>-0.045233</td>\n",
       "      <td>0.456895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">9996_273</th>\n",
       "      <th>82724002</th>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>0.733830</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.117389</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.634500</td>\n",
       "      <td>-0.046077</td>\n",
       "      <td>0.000014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38957146</th>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.032375</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.085699</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.045830</td>\n",
       "      <td>0.000007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5137608</th>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.167453</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.120247</td>\n",
       "      <td>-0.046283</td>\n",
       "      <td>0.000018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39393438</th>\n",
       "      <td>-0.679291</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.136425</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.204067</td>\n",
       "      <td>-0.046181</td>\n",
       "      <td>0.000008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45683368</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.51756</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.182602</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.014652</td>\n",
       "      <td>-0.046301</td>\n",
       "      <td>0.479649</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4000 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       prof_cat_exp_similarity  prof_cat_int_similarity  \\\n",
       "Job_ID   CANDIDATE_ID                                                     \n",
       "1014_514 23262979                     1.789496                 0.344139   \n",
       "         1482016                     -0.062095                -0.597146   \n",
       "         83352824                     0.555102                 0.344139   \n",
       "         54602123                     1.789496                 2.226710   \n",
       "         92641837                    -0.062095                -0.597146   \n",
       "...                                        ...                      ...   \n",
       "9996_273 82724002                    -0.679291                -0.597146   \n",
       "         38957146                    -0.679291                -0.032375   \n",
       "         5137608                     -0.679291                -0.597146   \n",
       "         39393438                    -0.679291                -0.597146   \n",
       "         45683368                     0.555102                 0.344139   \n",
       "\n",
       "                       hard_skills_similarity  soft_skills_similarity  \\\n",
       "Job_ID   CANDIDATE_ID                                                   \n",
       "1014_514 23262979                    0.887062                -0.51756   \n",
       "         1482016                    -0.564377                -0.51756   \n",
       "         83352824                    2.990909                -0.51756   \n",
       "         54602123                    0.887062                -0.51756   \n",
       "         92641837                   -0.564377                -0.51756   \n",
       "...                                       ...                     ...   \n",
       "9996_273 82724002                    0.733830                -0.51756   \n",
       "         38957146                   -0.564377                -0.51756   \n",
       "         5137608                    -0.564377                -0.51756   \n",
       "         39393438                   -0.564377                -0.51756   \n",
       "         45683368                   -0.564377                -0.51756   \n",
       "\n",
       "                       languages_similarity  cards_similarity  distance  \\\n",
       "Job_ID   CANDIDATE_ID                                                     \n",
       "1014_514 23262979                  2.776838         -0.519310 -0.092271   \n",
       "         1482016                   1.832306         -0.519310 -0.092271   \n",
       "         83352824                 -0.447997         -0.519310 -0.054167   \n",
       "         54602123                  2.776838         -0.519310 -0.055541   \n",
       "         92641837                  1.413862         -0.519310 -0.033892   \n",
       "...                                     ...               ...       ...   \n",
       "9996_273 82724002                 -0.447997          1.426400 -0.117389   \n",
       "         38957146                 -0.447997         -0.519310 -0.085699   \n",
       "         5137608                  -0.447997          2.232339 -0.167453   \n",
       "         39393438                 -0.447997          1.426400 -0.136425   \n",
       "         45683368                 -0.447997          2.232339 -0.182602   \n",
       "\n",
       "                       Education_similarity  overall_similarity  distance2  \\\n",
       "Job_ID   CANDIDATE_ID                                                        \n",
       "1014_514 23262979                 -0.022366            0.582359  -0.045888   \n",
       "         1482016                  -0.022366           -0.045517  -0.045888   \n",
       "         83352824                 -0.022366            0.582359  -0.045495   \n",
       "         54602123                 -0.022366            0.497398  -0.045512   \n",
       "         92641837                 -0.022366           -0.269762  -0.045233   \n",
       "...                                     ...                 ...        ...   \n",
       "9996_273 82724002                 -0.022366            0.634500  -0.046077   \n",
       "         38957146                 -0.022366           -0.903212  -0.045830   \n",
       "         5137608                  -0.022366           -0.120247  -0.046283   \n",
       "         39393438                 -0.022366            0.204067  -0.046181   \n",
       "         45683368                 -0.022366            1.014652  -0.046301   \n",
       "\n",
       "                       prediction  \n",
       "Job_ID   CANDIDATE_ID              \n",
       "1014_514 23262979        0.618107  \n",
       "         1482016         0.436729  \n",
       "         83352824        0.638060  \n",
       "         54602123        0.551775  \n",
       "         92641837        0.456895  \n",
       "...                           ...  \n",
       "9996_273 82724002        0.000014  \n",
       "         38957146        0.000007  \n",
       "         5137608         0.000018  \n",
       "         39393438        0.000008  \n",
       "         45683368        0.479649  \n",
       "\n",
       "[4000 rows x 11 columns]"
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "index_max = X_test.groupby(X_test.index.get_level_values(0))['prediction'].idxmax()\n",
    "X_test['Max'] = 0\n",
    "X_test.loc[index_max, 'Max'] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Job_ID\n",
       "1014_514     1\n",
       "101_663      1\n",
       "1026_1577    1\n",
       "102_3869     1\n",
       "1044_9332    1\n",
       "            ..\n",
       "9963_1394    1\n",
       "9965_4993    1\n",
       "9980_5692    1\n",
       "9982_5468    1\n",
       "9996_273     1\n",
       "Name: Max, Length: 800, dtype: int64"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.groupby(X_test.index.get_level_values(0))['Max'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1014_514</th>\n",
       "      <th>83352824</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.054167</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045495</td>\n",
       "      <td>0.638060</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101_663</th>\n",
       "      <th>9041634</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.166647</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046281</td>\n",
       "      <td>0.503321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026_1577</th>\n",
       "      <th>57238440</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>1.949590</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>0.856515</td>\n",
       "      <td>-0.188548</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.607861</td>\n",
       "      <td>-0.046303</td>\n",
       "      <td>0.536147</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102_3869</th>\n",
       "      <th>12929256</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>0.990016</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.180014</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.097548</td>\n",
       "      <td>-0.046300</td>\n",
       "      <td>0.586789</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1044_9332</th>\n",
       "      <th>70656680</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.173164</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046292</td>\n",
       "      <td>0.574584</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9963_1394</th>\n",
       "      <th>28215167</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.163185</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046274</td>\n",
       "      <td>0.503376</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9965_4993</th>\n",
       "      <th>71248428</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>0.743768</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.188548</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.087169</td>\n",
       "      <td>-0.046303</td>\n",
       "      <td>0.524201</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9980_5692</th>\n",
       "      <th>1899259</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.166162</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046281</td>\n",
       "      <td>0.511224</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9982_5468</th>\n",
       "      <th>82491379</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>1.488268</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>0.419353</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.204067</td>\n",
       "      <td>-0.029779</td>\n",
       "      <td>0.523786</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9996_273</th>\n",
       "      <th>45683368</th>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.182602</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.014652</td>\n",
       "      <td>-0.046301</td>\n",
       "      <td>0.479649</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>801 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        prof_cat_exp_similarity  prof_cat_int_similarity  \\\n",
       "Job_ID    CANDIDATE_ID                                                     \n",
       "1014_514  83352824                     0.555102                 0.344139   \n",
       "101_663   9041634                      0.555102                -0.597146   \n",
       "1026_1577 57238440                     1.789496                 0.344139   \n",
       "102_3869  12929256                     1.789496                -0.597146   \n",
       "1044_9332 70656680                     1.789496                -0.597146   \n",
       "...                                         ...                      ...   \n",
       "9963_1394 28215167                     0.555102                -0.597146   \n",
       "9965_4993 71248428                     0.555102                -0.597146   \n",
       "9980_5692 1899259                      1.789496                 0.344139   \n",
       "9982_5468 82491379                     0.555102                 0.344139   \n",
       "9996_273  45683368                     0.555102                 0.344139   \n",
       "\n",
       "                        hard_skills_similarity  soft_skills_similarity  \\\n",
       "Job_ID    CANDIDATE_ID                                                   \n",
       "1014_514  83352824                    2.990909               -0.517560   \n",
       "101_663   9041634                    -0.564377               -0.517560   \n",
       "1026_1577 57238440                    1.949590                2.302857   \n",
       "102_3869  12929256                   -0.564377                0.990016   \n",
       "1044_9332 70656680                   -0.564377               -0.517560   \n",
       "...                                        ...                     ...   \n",
       "9963_1394 28215167                   -0.564377               -0.517560   \n",
       "9965_4993 71248428                   -0.564377                0.743768   \n",
       "9980_5692 1899259                    -0.564377               -0.517560   \n",
       "9982_5468 82491379                    1.488268               -0.517560   \n",
       "9996_273  45683368                   -0.564377               -0.517560   \n",
       "\n",
       "                        languages_similarity  cards_similarity  distance  \\\n",
       "Job_ID    CANDIDATE_ID                                                     \n",
       "1014_514  83352824                 -0.447997         -0.519310 -0.054167   \n",
       "101_663   9041634                  -0.447997         -0.519310 -0.166647   \n",
       "1026_1577 57238440                 -0.447997          0.856515 -0.188548   \n",
       "102_3869  12929256                 -0.447997         -0.519310 -0.180014   \n",
       "1044_9332 70656680                 -0.447997         -0.519310 -0.173164   \n",
       "...                                      ...               ...       ...   \n",
       "9963_1394 28215167                 -0.447997         -0.519310 -0.163185   \n",
       "9965_4993 71248428                  2.776838         -0.519310 -0.188548   \n",
       "9980_5692 1899259                  -0.447997         -0.519310 -0.166162   \n",
       "9982_5468 82491379                 -0.447997         -0.519310  0.419353   \n",
       "9996_273  45683368                 -0.447997          2.232339 -0.182602   \n",
       "\n",
       "                        Education_similarity  overall_similarity  distance2  \\\n",
       "Job_ID    CANDIDATE_ID                                                        \n",
       "1014_514  83352824                 -0.022366            0.582359  -0.045495   \n",
       "101_663   9041634                  -0.022366           -0.903212  -0.046281   \n",
       "1026_1577 57238440                 -0.022366            1.607861  -0.046303   \n",
       "102_3869  12929256                 -0.022366           -0.097548  -0.046300   \n",
       "1044_9332 70656680                 -0.022366           -0.903212  -0.046292   \n",
       "...                                      ...                 ...        ...   \n",
       "9963_1394 28215167                 -0.022366           -0.903212  -0.046274   \n",
       "9965_4993 71248428                 -0.022366            0.087169  -0.046303   \n",
       "9980_5692 1899259                  -0.022366           -0.903212  -0.046281   \n",
       "9982_5468 82491379                 -0.022366            0.204067  -0.029779   \n",
       "9996_273  45683368                 -0.022366            1.014652  -0.046301   \n",
       "\n",
       "                        prediction  Max  \n",
       "Job_ID    CANDIDATE_ID                   \n",
       "1014_514  83352824        0.638060    1  \n",
       "101_663   9041634         0.503321    1  \n",
       "1026_1577 57238440        0.536147    1  \n",
       "102_3869  12929256        0.586789    1  \n",
       "1044_9332 70656680        0.574584    1  \n",
       "...                            ...  ...  \n",
       "9963_1394 28215167        0.503376    1  \n",
       "9965_4993 71248428        0.524201    1  \n",
       "9980_5692 1899259         0.511224    1  \n",
       "9982_5468 82491379        0.523786    1  \n",
       "9996_273  45683368        0.479649    1  \n",
       "\n",
       "[801 rows x 12 columns]"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_sub = X_test[X_test['Max'] == 1]\n",
    "X_test_sub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">6954_6795</th>\n",
       "      <th>43343785</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.108818</td>\n",
       "      <td>0.779395</td>\n",
       "      <td>1.875643</td>\n",
       "      <td>1.16442</td>\n",
       "      <td>-0.51931</td>\n",
       "      <td>0.242734</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.08989</td>\n",
       "      <td>-0.037986</td>\n",
       "      <td>0.575103</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43343785</th>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.108818</td>\n",
       "      <td>0.779395</td>\n",
       "      <td>1.875643</td>\n",
       "      <td>1.16442</td>\n",
       "      <td>-0.51931</td>\n",
       "      <td>0.242734</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.08989</td>\n",
       "      <td>-0.037986</td>\n",
       "      <td>0.575103</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        prof_cat_exp_similarity  prof_cat_int_similarity  \\\n",
       "Job_ID    CANDIDATE_ID                                                     \n",
       "6954_6795 43343785                     1.789496                 0.108818   \n",
       "          43343785                     1.789496                 0.108818   \n",
       "\n",
       "                        hard_skills_similarity  soft_skills_similarity  \\\n",
       "Job_ID    CANDIDATE_ID                                                   \n",
       "6954_6795 43343785                    0.779395                1.875643   \n",
       "          43343785                    0.779395                1.875643   \n",
       "\n",
       "                        languages_similarity  cards_similarity  distance  \\\n",
       "Job_ID    CANDIDATE_ID                                                     \n",
       "6954_6795 43343785                   1.16442          -0.51931  0.242734   \n",
       "          43343785                   1.16442          -0.51931  0.242734   \n",
       "\n",
       "                        Education_similarity  overall_similarity  distance2  \\\n",
       "Job_ID    CANDIDATE_ID                                                        \n",
       "6954_6795 43343785                 -0.022366             1.08989  -0.037986   \n",
       "          43343785                 -0.022366             1.08989  -0.037986   \n",
       "\n",
       "                        prediction  Max  \n",
       "Job_ID    CANDIDATE_ID                   \n",
       "6954_6795 43343785        0.575103    1  \n",
       "          43343785        0.575103    1  "
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_sub[X_test_sub.index.get_level_values(0) == '6954_6795']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [],
   "source": [
    "xx = X_test_sub.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1014_514</td>\n",
       "      <td>83352824</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.054167</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.582359</td>\n",
       "      <td>-0.045495</td>\n",
       "      <td>0.638060</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>101_663</td>\n",
       "      <td>9041634</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.166647</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046281</td>\n",
       "      <td>0.503321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1026_1577</td>\n",
       "      <td>57238440</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>1.949590</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>0.856515</td>\n",
       "      <td>-0.188548</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.607861</td>\n",
       "      <td>-0.046303</td>\n",
       "      <td>0.536147</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>102_3869</td>\n",
       "      <td>12929256</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>0.990016</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.180014</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.097548</td>\n",
       "      <td>-0.046300</td>\n",
       "      <td>0.586789</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1044_9332</td>\n",
       "      <td>70656680</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.173164</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046292</td>\n",
       "      <td>0.574584</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>9963_1394</td>\n",
       "      <td>28215167</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.163185</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046274</td>\n",
       "      <td>0.503376</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>9965_4993</td>\n",
       "      <td>71248428</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>0.743768</td>\n",
       "      <td>2.776838</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.188548</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.087169</td>\n",
       "      <td>-0.046303</td>\n",
       "      <td>0.524201</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>9980_5692</td>\n",
       "      <td>1899259</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.166162</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046281</td>\n",
       "      <td>0.511224</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>9982_5468</td>\n",
       "      <td>82491379</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>1.488268</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>0.419353</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.204067</td>\n",
       "      <td>-0.029779</td>\n",
       "      <td>0.523786</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>800</th>\n",
       "      <td>9996_273</td>\n",
       "      <td>45683368</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>0.344139</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.182602</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.014652</td>\n",
       "      <td>-0.046301</td>\n",
       "      <td>0.479649</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>801 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "0     1014_514      83352824                 0.555102   \n",
       "1      101_663       9041634                 0.555102   \n",
       "2    1026_1577      57238440                 1.789496   \n",
       "3     102_3869      12929256                 1.789496   \n",
       "4    1044_9332      70656680                 1.789496   \n",
       "..         ...           ...                      ...   \n",
       "796  9963_1394      28215167                 0.555102   \n",
       "797  9965_4993      71248428                 0.555102   \n",
       "798  9980_5692       1899259                 1.789496   \n",
       "799  9982_5468      82491379                 0.555102   \n",
       "800   9996_273      45683368                 0.555102   \n",
       "\n",
       "     prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "0                   0.344139                2.990909               -0.517560   \n",
       "1                  -0.597146               -0.564377               -0.517560   \n",
       "2                   0.344139                1.949590                2.302857   \n",
       "3                  -0.597146               -0.564377                0.990016   \n",
       "4                  -0.597146               -0.564377               -0.517560   \n",
       "..                       ...                     ...                     ...   \n",
       "796                -0.597146               -0.564377               -0.517560   \n",
       "797                -0.597146               -0.564377                0.743768   \n",
       "798                 0.344139               -0.564377               -0.517560   \n",
       "799                 0.344139                1.488268               -0.517560   \n",
       "800                 0.344139               -0.564377               -0.517560   \n",
       "\n",
       "     languages_similarity  cards_similarity  distance  Education_similarity  \\\n",
       "0               -0.447997         -0.519310 -0.054167             -0.022366   \n",
       "1               -0.447997         -0.519310 -0.166647             -0.022366   \n",
       "2               -0.447997          0.856515 -0.188548             -0.022366   \n",
       "3               -0.447997         -0.519310 -0.180014             -0.022366   \n",
       "4               -0.447997         -0.519310 -0.173164             -0.022366   \n",
       "..                    ...               ...       ...                   ...   \n",
       "796             -0.447997         -0.519310 -0.163185             -0.022366   \n",
       "797              2.776838         -0.519310 -0.188548             -0.022366   \n",
       "798             -0.447997         -0.519310 -0.166162             -0.022366   \n",
       "799             -0.447997         -0.519310  0.419353             -0.022366   \n",
       "800             -0.447997          2.232339 -0.182602             -0.022366   \n",
       "\n",
       "     overall_similarity  distance2  prediction  Max  \n",
       "0              0.582359  -0.045495    0.638060    1  \n",
       "1             -0.903212  -0.046281    0.503321    1  \n",
       "2              1.607861  -0.046303    0.536147    1  \n",
       "3             -0.097548  -0.046300    0.586789    1  \n",
       "4             -0.903212  -0.046292    0.574584    1  \n",
       "..                  ...        ...         ...  ...  \n",
       "796           -0.903212  -0.046274    0.503376    1  \n",
       "797            0.087169  -0.046303    0.524201    1  \n",
       "798           -0.903212  -0.046281    0.511224    1  \n",
       "799            0.204067  -0.029779    0.523786    1  \n",
       "800            1.014652  -0.046301    0.479649    1  \n",
       "\n",
       "[801 rows x 14 columns]"
      ]
     },
     "execution_count": 204,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [],
   "source": [
    "job_id_list_test = pd.read_csv('job_id_list_test.txt', header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2482_927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1310_19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8405_1436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6233_483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8105_8908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>795</th>\n",
       "      <td>1728_8760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>4803_9951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>4719_5184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>3999_3695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>7722_8343</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>800 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             0\n",
       "0     2482_927\n",
       "1      1310_19\n",
       "2    8405_1436\n",
       "3     6233_483\n",
       "4    8105_8908\n",
       "..         ...\n",
       "795  1728_8760\n",
       "796  4803_9951\n",
       "797  4719_5184\n",
       "798  3999_3695\n",
       "799  7722_8343\n",
       "\n",
       "[800 rows x 1 columns]"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "job_id_list_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2482_927</td>\n",
       "      <td>2482_927</td>\n",
       "      <td>64595902</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.135636</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>2.228646</td>\n",
       "      <td>-0.046178</td>\n",
       "      <td>0.703046</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1310_19</td>\n",
       "      <td>1310_19</td>\n",
       "      <td>4824920</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>1.832306</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.123622</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.202907</td>\n",
       "      <td>-0.046114</td>\n",
       "      <td>0.598643</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8405_1436</td>\n",
       "      <td>8405_1436</td>\n",
       "      <td>84114500</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>1.949590</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.168668</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.809057</td>\n",
       "      <td>-0.046285</td>\n",
       "      <td>0.499044</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6233_483</td>\n",
       "      <td>6233_483</td>\n",
       "      <td>55096946</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.814782</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.095272</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.045914</td>\n",
       "      <td>0.498017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8105_8908</td>\n",
       "      <td>8105_8908</td>\n",
       "      <td>25162353</td>\n",
       "      <td>0.143638</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>1.025595</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.175957</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.262166</td>\n",
       "      <td>-0.046296</td>\n",
       "      <td>0.543120</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>1728_8760</td>\n",
       "      <td>1728_8760</td>\n",
       "      <td>1999789</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>0.507582</td>\n",
       "      <td>2.141554</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.085031</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.296729</td>\n",
       "      <td>-0.045824</td>\n",
       "      <td>0.613596</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>4803_9951</td>\n",
       "      <td>4803_9951</td>\n",
       "      <td>11512982</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.145211</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046219</td>\n",
       "      <td>0.575629</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>4719_5184</td>\n",
       "      <td>4719_5184</td>\n",
       "      <td>79287820</td>\n",
       "      <td>0.143638</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>1.110808</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.170661</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.375364</td>\n",
       "      <td>-0.046289</td>\n",
       "      <td>0.472679</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>3999_3695</td>\n",
       "      <td>3999_3695</td>\n",
       "      <td>24714973</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.170144</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.757707</td>\n",
       "      <td>-0.046288</td>\n",
       "      <td>0.502295</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>800</th>\n",
       "      <td>7722_8343</td>\n",
       "      <td>7722_8343</td>\n",
       "      <td>29815140</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.134225</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046171</td>\n",
       "      <td>0.576040</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>801 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             0     Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "0     2482_927   2482_927      64595902                 1.789496   \n",
       "1      1310_19    1310_19       4824920                 1.789496   \n",
       "2    8405_1436  8405_1436      84114500                 0.555102   \n",
       "3     6233_483   6233_483      55096946                 1.789496   \n",
       "4    8105_8908  8105_8908      25162353                 0.143638   \n",
       "..         ...        ...           ...                      ...   \n",
       "796  1728_8760  1728_8760       1999789                 1.789496   \n",
       "797  4803_9951  4803_9951      11512982                 1.789496   \n",
       "798  4719_5184  4719_5184      79287820                 0.143638   \n",
       "799  3999_3695  3999_3695      24714973                 0.555102   \n",
       "800  7722_8343  7722_8343      29815140                 1.789496   \n",
       "\n",
       "     prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "0                  -0.597146                2.990909                2.302857   \n",
       "1                  -0.597146               -0.564377               -0.517560   \n",
       "2                  -0.597146                1.949590                2.302857   \n",
       "3                   0.814782               -0.564377               -0.517560   \n",
       "4                  -0.597146                1.025595               -0.517560   \n",
       "..                       ...                     ...                     ...   \n",
       "796                -0.597146                0.507582                2.141554   \n",
       "797                -0.597146               -0.564377               -0.517560   \n",
       "798                -0.597146               -0.564377                1.110808   \n",
       "799                -0.597146               -0.564377               -0.517560   \n",
       "800                -0.597146               -0.564377               -0.517560   \n",
       "\n",
       "     languages_similarity  cards_similarity  distance  Education_similarity  \\\n",
       "0               -0.447997         -0.519310 -0.135636             -0.022366   \n",
       "1                1.832306         -0.519310 -0.123622             -0.022366   \n",
       "2               -0.447997         -0.519310 -0.168668             -0.022366   \n",
       "3               -0.447997         -0.519310 -0.095272             -0.022366   \n",
       "4                1.413862         -0.519310 -0.175957             -0.022366   \n",
       "..                    ...               ...       ...                   ...   \n",
       "796              1.413862          1.426400 -0.085031             -0.022366   \n",
       "797             -0.447997         -0.519310 -0.145211             -0.022366   \n",
       "798             -0.447997         -0.519310 -0.170661             -0.022366   \n",
       "799             -0.447997          2.232339 -0.170144             -0.022366   \n",
       "800             -0.447997         -0.519310 -0.134225             -0.022366   \n",
       "\n",
       "     overall_similarity  distance2  prediction  Max  \n",
       "0              2.228646  -0.046178    0.703046    1  \n",
       "1             -0.202907  -0.046114    0.598643    1  \n",
       "2              1.809057  -0.046285    0.499044    1  \n",
       "3             -0.903212  -0.045914    0.498017    1  \n",
       "4              0.262166  -0.046296    0.543120    1  \n",
       "..                  ...        ...         ...  ...  \n",
       "796            1.296729  -0.045824    0.613596    1  \n",
       "797           -0.903212  -0.046219    0.575629    1  \n",
       "798            0.375364  -0.046289    0.472679    1  \n",
       "799            0.757707  -0.046288    0.502295    1  \n",
       "800           -0.903212  -0.046171    0.576040    1  \n",
       "\n",
       "[801 rows x 15 columns]"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final = pd.merge(job_id_list_test, xx, left_on=0, right_on='Job_ID', how='left')\n",
    "final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>6954_6795</td>\n",
       "      <td>6954_6795</td>\n",
       "      <td>43343785</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.108818</td>\n",
       "      <td>0.779395</td>\n",
       "      <td>1.875643</td>\n",
       "      <td>1.16442</td>\n",
       "      <td>-0.51931</td>\n",
       "      <td>0.242734</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.08989</td>\n",
       "      <td>-0.037986</td>\n",
       "      <td>0.575103</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             0     Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "146  6954_6795  6954_6795      43343785                 1.789496   \n",
       "\n",
       "     prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "146                 0.108818                0.779395                1.875643   \n",
       "\n",
       "     languages_similarity  cards_similarity  distance  Education_similarity  \\\n",
       "146               1.16442          -0.51931  0.242734             -0.022366   \n",
       "\n",
       "     overall_similarity  distance2  prediction  Max  \n",
       "146             1.08989  -0.037986    0.575103    1  "
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final[final.duplicated(subset=[0])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [],
   "source": [
    "ff = final.drop_duplicates(subset=[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>Job_ID</th>\n",
       "      <th>CANDIDATE_ID</th>\n",
       "      <th>prof_cat_exp_similarity</th>\n",
       "      <th>prof_cat_int_similarity</th>\n",
       "      <th>hard_skills_similarity</th>\n",
       "      <th>soft_skills_similarity</th>\n",
       "      <th>languages_similarity</th>\n",
       "      <th>cards_similarity</th>\n",
       "      <th>distance</th>\n",
       "      <th>Education_similarity</th>\n",
       "      <th>overall_similarity</th>\n",
       "      <th>distance2</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2482_927</td>\n",
       "      <td>2482_927</td>\n",
       "      <td>64595902</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>2.990909</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.135636</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>2.228646</td>\n",
       "      <td>-0.046178</td>\n",
       "      <td>0.703046</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1310_19</td>\n",
       "      <td>1310_19</td>\n",
       "      <td>4824920</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>1.832306</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.123622</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.202907</td>\n",
       "      <td>-0.046114</td>\n",
       "      <td>0.598643</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8405_1436</td>\n",
       "      <td>8405_1436</td>\n",
       "      <td>84114500</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>1.949590</td>\n",
       "      <td>2.302857</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.168668</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.809057</td>\n",
       "      <td>-0.046285</td>\n",
       "      <td>0.499044</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6233_483</td>\n",
       "      <td>6233_483</td>\n",
       "      <td>55096946</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>0.814782</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.095272</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.045914</td>\n",
       "      <td>0.498017</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8105_8908</td>\n",
       "      <td>8105_8908</td>\n",
       "      <td>25162353</td>\n",
       "      <td>0.143638</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>1.025595</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.175957</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.262166</td>\n",
       "      <td>-0.046296</td>\n",
       "      <td>0.543120</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>796</th>\n",
       "      <td>1728_8760</td>\n",
       "      <td>1728_8760</td>\n",
       "      <td>1999789</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>0.507582</td>\n",
       "      <td>2.141554</td>\n",
       "      <td>1.413862</td>\n",
       "      <td>1.426400</td>\n",
       "      <td>-0.085031</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>1.296729</td>\n",
       "      <td>-0.045824</td>\n",
       "      <td>0.613596</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>797</th>\n",
       "      <td>4803_9951</td>\n",
       "      <td>4803_9951</td>\n",
       "      <td>11512982</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.145211</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046219</td>\n",
       "      <td>0.575629</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>798</th>\n",
       "      <td>4719_5184</td>\n",
       "      <td>4719_5184</td>\n",
       "      <td>79287820</td>\n",
       "      <td>0.143638</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>1.110808</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.170661</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.375364</td>\n",
       "      <td>-0.046289</td>\n",
       "      <td>0.472679</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799</th>\n",
       "      <td>3999_3695</td>\n",
       "      <td>3999_3695</td>\n",
       "      <td>24714973</td>\n",
       "      <td>0.555102</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>2.232339</td>\n",
       "      <td>-0.170144</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>0.757707</td>\n",
       "      <td>-0.046288</td>\n",
       "      <td>0.502295</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>800</th>\n",
       "      <td>7722_8343</td>\n",
       "      <td>7722_8343</td>\n",
       "      <td>29815140</td>\n",
       "      <td>1.789496</td>\n",
       "      <td>-0.597146</td>\n",
       "      <td>-0.564377</td>\n",
       "      <td>-0.517560</td>\n",
       "      <td>-0.447997</td>\n",
       "      <td>-0.519310</td>\n",
       "      <td>-0.134225</td>\n",
       "      <td>-0.022366</td>\n",
       "      <td>-0.903212</td>\n",
       "      <td>-0.046171</td>\n",
       "      <td>0.576040</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>800 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             0     Job_ID  CANDIDATE_ID  prof_cat_exp_similarity  \\\n",
       "0     2482_927   2482_927      64595902                 1.789496   \n",
       "1      1310_19    1310_19       4824920                 1.789496   \n",
       "2    8405_1436  8405_1436      84114500                 0.555102   \n",
       "3     6233_483   6233_483      55096946                 1.789496   \n",
       "4    8105_8908  8105_8908      25162353                 0.143638   \n",
       "..         ...        ...           ...                      ...   \n",
       "796  1728_8760  1728_8760       1999789                 1.789496   \n",
       "797  4803_9951  4803_9951      11512982                 1.789496   \n",
       "798  4719_5184  4719_5184      79287820                 0.143638   \n",
       "799  3999_3695  3999_3695      24714973                 0.555102   \n",
       "800  7722_8343  7722_8343      29815140                 1.789496   \n",
       "\n",
       "     prof_cat_int_similarity  hard_skills_similarity  soft_skills_similarity  \\\n",
       "0                  -0.597146                2.990909                2.302857   \n",
       "1                  -0.597146               -0.564377               -0.517560   \n",
       "2                  -0.597146                1.949590                2.302857   \n",
       "3                   0.814782               -0.564377               -0.517560   \n",
       "4                  -0.597146                1.025595               -0.517560   \n",
       "..                       ...                     ...                     ...   \n",
       "796                -0.597146                0.507582                2.141554   \n",
       "797                -0.597146               -0.564377               -0.517560   \n",
       "798                -0.597146               -0.564377                1.110808   \n",
       "799                -0.597146               -0.564377               -0.517560   \n",
       "800                -0.597146               -0.564377               -0.517560   \n",
       "\n",
       "     languages_similarity  cards_similarity  distance  Education_similarity  \\\n",
       "0               -0.447997         -0.519310 -0.135636             -0.022366   \n",
       "1                1.832306         -0.519310 -0.123622             -0.022366   \n",
       "2               -0.447997         -0.519310 -0.168668             -0.022366   \n",
       "3               -0.447997         -0.519310 -0.095272             -0.022366   \n",
       "4                1.413862         -0.519310 -0.175957             -0.022366   \n",
       "..                    ...               ...       ...                   ...   \n",
       "796              1.413862          1.426400 -0.085031             -0.022366   \n",
       "797             -0.447997         -0.519310 -0.145211             -0.022366   \n",
       "798             -0.447997         -0.519310 -0.170661             -0.022366   \n",
       "799             -0.447997          2.232339 -0.170144             -0.022366   \n",
       "800             -0.447997         -0.519310 -0.134225             -0.022366   \n",
       "\n",
       "     overall_similarity  distance2  prediction  Max  \n",
       "0              2.228646  -0.046178    0.703046    1  \n",
       "1             -0.202907  -0.046114    0.598643    1  \n",
       "2              1.809057  -0.046285    0.499044    1  \n",
       "3             -0.903212  -0.045914    0.498017    1  \n",
       "4              0.262166  -0.046296    0.543120    1  \n",
       "..                  ...        ...         ...  ...  \n",
       "796            1.296729  -0.045824    0.613596    1  \n",
       "797           -0.903212  -0.046219    0.575629    1  \n",
       "798            0.375364  -0.046289    0.472679    1  \n",
       "799            0.757707  -0.046288    0.502295    1  \n",
       "800           -0.903212  -0.046171    0.576040    1  \n",
       "\n",
       "[800 rows x 15 columns]"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ff"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}