part2_submission_Training_Evaluation_Credit_risk.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.metrics import accuracy_score,confusion_matrix,ConfusionMatrixDisplay,roc_curve, auc,roc_auc_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Features: \n",
      " fcredit_amount, checking_status, age, duration, purpose, credit_history, employment, savings_status, property_magnitude, residence_since, installment_commitment, personal_status, job, other_payment_plans, housing, existing_credits, class\n",
      "17\n",
      "\n",
      "Classification Data Type: \n",
      " fcredit_amount             float64\n",
      "checking_status             int64\n",
      "age                       float64\n",
      "duration                  float64\n",
      "purpose                     int64\n",
      "credit_history              int64\n",
      "employment                  int64\n",
      "savings_status              int64\n",
      "property_magnitude          int64\n",
      "residence_since           float64\n",
      "installment_commitment    float64\n",
      "personal_status             int64\n",
      "job                         int64\n",
      "other_payment_plans         int64\n",
      "housing                     int64\n",
      "existing_credits          float64\n",
      "class                       int64\n",
      "dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Load Classification Data\n",
    "classification_data_df = pd.read_csv(\"./data/credit_classification_encode_fs_n_b89.csv\")\n",
    "classification_feature_list = classification_data_df.columns\n",
    "# Features: fcredit_amount, checking_status, age, duration, purpose, credit_history, employment, savings_status, property_magnitude, residence_since, installment_commitment, personal_status, job, other_payment_plans, housing, existing_credits, class\n",
    "print(f\"Classification Features: \\n f{', '.join(classification_feature_list)}\")\n",
    "print(len(classification_feature_list))\n",
    "print()\n",
    "#Our target will be 'Class', we will using 1 present good and 0 present bad\n",
    "classification_data_type = classification_data_df.dtypes \n",
    "print(f\"Classification Data Type: \\n f{classification_data_type}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data we first feed has 17 columns and all numeric type:. \n",
    "Our target is ‘Class: good/bad’(discrete data)\n",
    "I have encoded categorical type using label encoder and ordinal encoder, I have done normalization to prevent larger scaled features dominant the result, I also used feature selection to remove unimportant data (the threshold for this stage is small as I don’t want to filter out potential useful information), and I resampled the target class as the target class is unbalanced.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split Classification dataset into a training set and a test set\n",
    "Classification_X = classification_data_df.drop('class',axis=1)\n",
    "Classification_y = classification_data_df[\"class\"]\n",
    "Classification_X_train,Classification_X_test,Classification_y_train,Classification_y_test = train_test_split(Classification_X,Classification_y,test_size=0.2,random_state=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this training and evaluation process, I have split the data into training_set (80%) and test_set(20%). As the model should work well not only within the current given data, but also robust to potential unknown data in real world. If we use all the current data we may have relatively good result, but without splitting and compare our model performance, the model may only perform well with given data and not robust to unknown real world result. This is the case where the model overfitting. And if the model even not work well with current given data, which means it is likely to underfit. In another word, the model did not find the real insights. We try to minimize the possibility to be overfit or underfit and we try to make the model more robust for future use. \n",
    "Using split data could help to justify the test design. It is more reliable as the data from the same dataset should follow the same distribution. (Just be mindful not overlap training and testing data in case of introducing overfitting)\n",
    "Additionally, training set should larger than the test set. With sufficient data, the training model will be more representative and robust. The test set is just used to verify the training set build model performance.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Choose an algorithm for Classification \n",
    "# Random Forrest\n",
    "# Logistic regression\n",
    "# Random Forrest\n",
    "\n",
    "# This study is a supervised learning study, as all the data points are labelled, use the data points as inputs to build a model to get a desired output. And target class is discrete data, which is classification question. \n",
    "# Regression model used to predicts real numbers. Rather than saying whether the object is a mushroom or not, it mainly focuses on determining the probability of the object being/not being a mushroom. \n",
    "\n",
    "\n",
    "# Random Forrest and Gradient Boosting are Ensembles method. It is similar to the tree algorithm and could deal with both classification and regression problem. Ensembles combines a set of classifiers and combine their decisions in some way to get the result (voting, stacking and etc). It could handle errors or even correct errors\n",
    "# But there is some constraints : each classier needs to be better than random (otherwise the combination may result worse); and each classifier needs to disagree with each other(they will infer each other)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train and Test a Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters for RF model {'n_estimators': 86, 'min_samples_split': 4, 'min_samples_leaf': 4, 'max_features': 6, 'max_depth': None, 'criterion': 'entropy', 'bootstrap': True}\n",
      "Accuracy: 79.64285714285714%\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "# Random Forrest:\n",
    "# Parameter set for RF model\n",
    "rf_param_dict = {\n",
    "    \"n_estimators\": range(10, 100), #It represents the number of trees in the random forest.\n",
    "    \"max_depth\": [3, None], #It controls the maximum depth of the individual decision trees in the forest.\n",
    "    \"max_features\": range(1, 11), #It determines the maximum number of features to consider for splitting at each node.\n",
    "    \"min_samples_split\": range(2, 11), #It represents the minimum number of samples required to split an internal node.\n",
    "    \"min_samples_leaf\": range(1, 11), #It represents the minimum number of samples required to be at a leaf node.\n",
    "    \"bootstrap\": [True, False], #It determines whether bootstrap samples are used when building trees.\n",
    "    \"criterion\": [\"gini\", \"entropy\"] #It represents the function to measure the quality of a split.\n",
    "}\n",
    "classification_rf_model = RandomForestClassifier()\n",
    "\n",
    "# using random search to find out the best parameter setting for RF\n",
    "random_search_rf = RandomizedSearchCV(classification_rf_model, param_distributions = rf_param_dict, n_iter = 20, cv = 5)\n",
    "random_search_rf.fit(Classification_X_train, Classification_y_train)\n",
    "print(f\"Best parameters for RF model {random_search_rf.best_params_}\")\n",
    "\n",
    "classification_model_rf = random_search_rf.best_estimator_\n",
    "# train the model using training set\n",
    "classification_model_rf.fit(Classification_X_train, Classification_y_train)\n",
    "# using model to do perdiction\n",
    "classification_model_rf_predict = classification_model_rf.predict(Classification_X_test)\n",
    "# calculate  accuracy of the model \n",
    "classification_model_rf_accuracy = accuracy_score(Classification_y_test, classification_model_rf_predict)\n",
    "\n",
    "print(f\"Accuracy: {classification_model_rf_accuracy * 100}%\")\n",
    "pickle.dump(classification_model_rf, open(\"credit_random_forrest.pkl\", \"wb\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\linear_model\\_sag.py:350: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters for LR model {'solver': 'liblinear', 'penalty': 'l1', 'C': 1}\n",
      "Accuracy: 72.85714285714285%\n"
     ]
    }
   ],
   "source": [
    "#Logistic Regression\n",
    "#Parameter set for LR model\n",
    "lr_param_dist = {\n",
    "    'C': [0.001, 0.01, 0.1, 1, 10, 100],  # Inverse of regularization strength\n",
    "    'penalty': ['l1', 'l2'],              # Regularization type\n",
    "    'solver': ['liblinear', 'saga']       # Solver type\n",
    "}\n",
    "classification_lr_model = LogisticRegression()\n",
    "lr_model_random_search = RandomizedSearchCV(classification_lr_model, param_distributions = lr_param_dist, n_iter = 20, cv = 5)\n",
    "# train the model using training set\n",
    "lr_model_random_search.fit(Classification_X_train, Classification_y_train)\n",
    "print(f\"Best parameters for LR model {lr_model_random_search.best_params_}\")\n",
    "classification_model_lr = lr_model_random_search.best_estimator_\n",
    "# using model to do perdiction\n",
    "classification_model_lr.fit(Classification_X_train, Classification_y_train)\n",
    "# using model to do perdiction\n",
    "classification_model_lr_predict = classification_model_lr.predict(Classification_X_test)\n",
    "# calculate  accuracy of the model \n",
    "\n",
    "classification_model_lr_accuracy = accuracy_score(np.array(Classification_y_test), classification_model_lr_predict)\n",
    "\n",
    "print(f\"Accuracy: {classification_model_lr_accuracy * 100}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\model_selection\\_validation.py:425: FitFailedWarning: \n",
      "20 fits failed out of a total of 100.\n",
      "The score on these train-test partitions for these parameters will be set to nan.\n",
      "If these failures are not expected, you can try to debug them by setting error_score='raise'.\n",
      "\n",
      "Below are more details about the failures:\n",
      "--------------------------------------------------------------------------------\n",
      "20 fits failed with the following error:\n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\model_selection\\_validation.py\", line 732, in _fit_and_score\n",
      "    estimator.fit(X_train, y_train, **fit_params)\n",
      "  File \"c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\base.py\", line 1144, in wrapper\n",
      "    estimator._validate_params()\n",
      "  File \"c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\base.py\", line 637, in _validate_params\n",
      "    validate_parameter_constraints(\n",
      "  File \"c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\utils\\_param_validation.py\", line 95, in validate_parameter_constraints\n",
      "    raise InvalidParameterError(\n",
      "sklearn.utils._param_validation.InvalidParameterError: The 'max_features' parameter of GradientBoostingClassifier must be an int in the range [1, inf), a float in the range (0.0, 1.0], a str among {'sqrt', 'log2'} or None. Got 'auto' instead.\n",
      "\n",
      "  warnings.warn(some_fits_failed_message, FitFailedWarning)\n",
      "c:\\Users\\Kitty\\Desktop\\final\\venv-msa\\Lib\\site-packages\\sklearn\\model_selection\\_search.py:976: UserWarning: One or more of the test scores are non-finite: [0.8125     0.76339286 0.78035714 0.8        0.79375    0.76428571\n",
      " 0.79910714 0.79017857 0.76875    0.79642857 0.74375           nan\n",
      "        nan 0.79107143 0.76160714 0.79285714        nan        nan\n",
      " 0.7875     0.81428571]\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters for GB model {'n_estimators': 69, 'min_samples_split': 8, 'min_samples_leaf': 9, 'max_features': 'log2', 'max_depth': None, 'learning_rate': 0.1}\n",
      "Accuracy: 83.21428571428572%\n"
     ]
    }
   ],
   "source": [
    "# Gradient boosting\n",
    "#Parameter set for GB model\n",
    "gb_param_dist = {\n",
    "    'n_estimators': np.arange(10, 100),\n",
    "    'learning_rate': [0.001, 0.01, 0.1, 0.2],\n",
    "    'max_depth': [3, 4, 5, None],\n",
    "    'min_samples_split': np.arange(2, 11),\n",
    "    'min_samples_leaf': np.arange(1, 11),\n",
    "    'max_features': ['auto', 'sqrt', 'log2', None]\n",
    "}\n",
    "\n",
    "\n",
    "classification_gb_model = GradientBoostingClassifier()\n",
    "gb_model_random_search = RandomizedSearchCV(classification_gb_model, param_distributions = gb_param_dist, n_iter = 20, cv = 5)\n",
    "# train the model using training set\n",
    "gb_model_random_search.fit(Classification_X_train, Classification_y_train)\n",
    "print(f\"Best parameters for GB model {gb_model_random_search.best_params_}\")\n",
    "\n",
    "classification_model_gb = gb_model_random_search.best_estimator_\n",
    "# using model to do perdiction\n",
    "classification_model_gb.fit(Classification_X_train, Classification_y_train)\n",
    "# using model to do perdiction\n",
    "classification_model_gb_predict = classification_model_gb.predict(Classification_X_test)\n",
    "# calculate  accuracy of the model \n",
    "classification_model_gb_accuracy = accuracy_score(Classification_y_test, classification_model_gb_predict)\n",
    "print(f\"Accuracy: {classification_model_gb_accuracy * 100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate the model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall: 0.8085106382978723\n",
      "Precision: 0.7916666666666666\n",
      "F-measure: 0.7999999999999999\n",
      "Confusion Matrix for Random Forrest:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1 AUC: 0.8161\n",
      "Fold 2 AUC: 0.7973\n",
      "Fold 3 AUC: 0.8076\n",
      "Fold 4 AUC: 0.8341\n",
      "Fold 5 AUC: 0.9043\n",
      "Fold 6 AUC: 0.9857\n",
      "Fold 7 AUC: 0.9578\n",
      "Fold 8 AUC: 0.9561\n",
      "Fold 9 AUC: 0.9241\n",
      "Fold 10 AUC: 0.9633\n",
      "Mean AUC: 0.8946\n"
     ]
    }
   ],
   "source": [
    "# Confusion Matrix for Random Forrest\n",
    "rf_confusion_matrix = confusion_matrix(Classification_y_test, classification_model_rf_predict)\n",
    "tn, fp, fn, tp = rf_confusion_matrix.ravel()\n",
    "recall = tp /(tp + fn)\n",
    "precision = tp / (tp + fp)\n",
    "f_measure = ((2* recall)*precision)/(recall + precision)\n",
    "print(f\"Recall: {recall}\")\n",
    "print(f\"Precision: {precision}\")\n",
    "print(f\"F-measure: {f_measure}\")\n",
    "print(\"Confusion Matrix for Random Forrest:\")\n",
    "\n",
    "rf_matrix_display = ConfusionMatrixDisplay(rf_confusion_matrix).plot()\n",
    "\n",
    "\n",
    "#ROC Cure and AUC and 10 cross validation\n",
    "cv = StratifiedKFold(n_splits=10)\n",
    "\n",
    "roc_curves = []\n",
    "auc_scores = []\n",
    "\n",
    "\n",
    "for train_idx, test_idx in cv.split(Classification_X,Classification_y):\n",
    "    X_train, X_test = Classification_X.iloc[train_idx], Classification_X.iloc[test_idx]\n",
    "    y_train, y_test = Classification_y.iloc[train_idx], Classification_y.iloc[test_idx]\n",
    "    classification_model_rf.fit(X_train, y_train)\n",
    "    y_score = classification_model_rf.predict_proba(X_test)[:, 1]\n",
    "    fpr, tpr, _ = roc_curve(y_test, y_score)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    roc_curves.append((fpr, tpr, roc_auc))\n",
    "    auc_scores.append(roc_auc)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "for fpr, tpr, roc_auc in roc_curves:\n",
    "    plt.plot(fpr, tpr, lw=1, alpha=0.7)\n",
    "mean_fpr = np.linspace(0, 1, 100)\n",
    "mean_tpr = np.mean([np.interp(mean_fpr, fpr, tpr) for fpr, tpr, _ in roc_curves], axis=0)\n",
    "plt.plot(mean_fpr, mean_tpr, color='b', label='Mean ROC')\n",
    "plt.plot([0, 1], [0, 1], color='gray', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()\n",
    "for fold, auc_score in enumerate(auc_scores, start=1):\n",
    "    print(f\"Fold {fold} AUC: {auc_score:.4f}\")\n",
    "\n",
    "mean_auc = np.mean(auc_scores)\n",
    "print(f\"Mean AUC: {mean_auc:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall: 0.7588652482269503\n",
      "Precision: 0.7181208053691275\n",
      "F-measure: 0.7379310344827585\n",
      "Confusion Matrix for Logistic Regression:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1 AUC: 0.7049\n",
      "Fold 2 AUC: 0.7233\n",
      "Fold 3 AUC: 0.7143\n",
      "Fold 4 AUC: 0.6967\n",
      "Fold 5 AUC: 0.8120\n",
      "Fold 6 AUC: 0.8898\n",
      "Fold 7 AUC: 0.7878\n",
      "Fold 8 AUC: 0.8955\n",
      "Fold 9 AUC: 0.7969\n",
      "Fold 10 AUC: 0.8057\n",
      "Mean AUC: 0.7827\n"
     ]
    }
   ],
   "source": [
    "# Confusion Matrix for Logistic Regression\n",
    "lr_confusion_matrix = confusion_matrix(Classification_y_test, classification_model_lr_predict)\n",
    "tn, fp, fn, tp = lr_confusion_matrix.ravel()\n",
    "recall = tp /(tp + fn)\n",
    "precision = tp / (tp + fp)\n",
    "f_measure = ((2* recall)*precision)/(recall + precision)\n",
    "print(f\"Recall: {recall}\")\n",
    "print(f\"Precision: {precision}\")\n",
    "print(f\"F-measure: {f_measure}\")\n",
    "print(\"Confusion Matrix for Logistic Regression:\")\n",
    "\n",
    "\n",
    "\n",
    "ConfusionMatrixDisplay(lr_confusion_matrix).plot()\n",
    "\n",
    "#ROC Cure and AUC and 10 cross validation\n",
    "cv = StratifiedKFold(n_splits=10)\n",
    "\n",
    "roc_curves = []\n",
    "auc_scores = []\n",
    "\n",
    "\n",
    "for train_idx, test_idx in cv.split(Classification_X,Classification_y):\n",
    "    X_train, X_test = Classification_X.iloc[train_idx], Classification_X.iloc[test_idx]\n",
    "    y_train, y_test = Classification_y.iloc[train_idx], Classification_y.iloc[test_idx]\n",
    "    classification_model_lr.fit(X_train, y_train)\n",
    "    y_score = classification_model_lr.predict_proba(X_test)[:, 1]\n",
    "    fpr, tpr, _ = roc_curve(y_test, y_score)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    roc_curves.append((fpr, tpr, roc_auc))\n",
    "    auc_scores.append(roc_auc)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "for fpr, tpr, roc_auc in roc_curves:\n",
    "    plt.plot(fpr, tpr, lw=1, alpha=0.7)\n",
    "mean_fpr = np.linspace(0, 1, 100)\n",
    "mean_tpr = np.mean([np.interp(mean_fpr, fpr, tpr) for fpr, tpr, _ in roc_curves], axis=0)\n",
    "plt.plot(mean_fpr, mean_tpr, color='b', label='Mean ROC')\n",
    "plt.plot([0, 1], [0, 1], color='gray', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()\n",
    "for fold, auc_score in enumerate(auc_scores, start=1):\n",
    "    print(f\"Fold {fold} AUC: {auc_score:.4f}\")\n",
    "mean_auc = np.mean(auc_scores)\n",
    "print(f\"Mean AUC: {mean_auc:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall: 0.8581560283687943\n",
      "Precision: 0.8175675675675675\n",
      "F-measure: 0.8373702422145328\n",
      "Confusion Matrix for Gradient Boosting:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAGyCAYAAADj3G12AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAyMUlEQVR4nO3deXhU9dn/8c8kIQshC0HJogGjIItiUFAaV9AI4gYFa7HYRkSoShRIEaE1rEIUNxpEcIXSH7i0Fh6hLT4IyiKIEsTHJUaWKGFJUGPIAlmYOb8/UkanLGZyJpnMOe/XdZ3r6pxt7rFc3Nz393vO12EYhiEAAGBZQf4OAAAANC2SPQAAFkeyBwDA4kj2AABYHMkeAACLI9kDAGBxJHsAACyOZA8AgMWR7AEAsLgQfwdghsvl0oEDBxQVFSWHw+HvcAAAXjIMQxUVFUpKSlJQUNPVn9XV1aqtrTV9n9DQUIWHh/sgomZmBLCioiJDEhsbGxtbgG9FRUVNliuOHj1qJLQP9kmcCQkJxtGjRxv0vevXrzduvvlmIzEx0ZBkLF++3H2strbWmDhxonHhhRcarVu3NhITE43f/va3xv79+z3u8f333xu/+c1vjKioKCMmJsa4++67jYqKCq//GwR0ZR8VFSVJWr45WZFtGJGANc25vr+/QwCazDFXrd779i/uv8+bQm1trYoPOfVN3jmKjmp8riivcKljr69VW1vboOq+qqpKqampuvvuuzVkyBCPY0eOHNH27duVnZ2t1NRU/fDDDxo7dqxuvfVWbdu2zX3e8OHDdfDgQa1Zs0Z1dXUaMWKERo8erWXLlnkVe0An++Ot+8g2QYo08X8g0JKFBIX6OwSgyTXHUGybKIfaRDX+e1zy7tqBAwdq4MCBJz0WExOjNWvWeOx79tlnddlll2nv3r3q0KGD8vPztXr1an300Ufq3bu3JGnevHm68cYb9eSTTyopKanBsZAhAQC24DRcpjdJKi8v99hqamp8Et/hw4flcDgUGxsrSdqyZYtiY2PdiV6S0tPTFRQUpK1bt3p1b5I9AMAWXDJMb5KUnJysmJgY95aTk2M6turqaj388MO64447FB0dLUkqLi5W+/btPc4LCQlRXFyciouLvbp/QLfxAQBobkVFRe6ELElhYWGm7ldXV6fbb79dhmFowYIFZsM7KZI9AMAWXHLJZfJ6SYqOjvZI9mYcT/TffPON1q1b53HfhIQEHTp0yOP8Y8eOqbS0VAkJCV59D218AIAtOA3D9OZLxxP9zp079c4776hdu3Yex9PS0lRWVqa8vDz3vnXr1snlcqlPnz5efReVPQAATaCyslK7du1yfy4sLNSOHTsUFxenxMRE3Xbbbdq+fbtWrVolp9PpHoePi4tTaGiounXrphtuuEGjRo3SwoULVVdXp8zMTA0bNsyrmfgSyR4AYBM/nWTX2Ou9sW3bNvXr18/9OSsrS5KUkZGhadOm6a233pIk9ezZ0+O6d999V3379pUkLV26VJmZmbruuusUFBSkoUOHKjc31+vYSfYAAFtwyZCzGZN93759ZZym9X+6Y8fFxcV5/QKdk2HMHgAAi6OyBwDYQnO38VsSkj0AwBbMzqj39Wz85kQbHwAAi6OyBwDYgus/m5nrAxXJHgBgC06Ts/HNXOtvJHsAgC04jfrNzPWBijF7AAAsjsoeAGALjNkDAGBxLjnklMPU9YGKNj4AABZHZQ8AsAWXUb+ZuT5QkewBALbgNNnGN3Otv9HGBwDA4qjsAQC2YOfKnmQPALAFl+GQyzAxG9/Etf5GGx8AAIujsgcA2AJtfAAALM6pIDlNNLSdPoyluZHsAQC2YJgcszcYswcAAC0VlT0AwBYYswcAwOKcRpCchokx+wB+XS5tfAAALI7KHgBgCy455DJR47oUuKU9yR4AYAt2HrOnjQ8AgMVR2QMAbMH8BD3a+AAAtGj1Y/YmFsKhjQ8AAFoqKnsAgC24TL4bn9n4AAC0cIzZAwBgcS4F2fY5e8bsAQCwOCp7AIAtOA2HnCaWqTVzrb+R7AEAtuA0OUHPSRsfAAC0VFT2AABbcBlBcpmYje9iNj4AAC0bbXwAAGBZVPYAAFtwydyMepfvQml2JHsAgC2Yf6lO4DbDAzdyAADQIFT2AABbMP9u/MCtj0n2AABbsPN69iR7AIAt2LmyD9zIAQBAg1DZAwBswfxLdQK3PibZAwBswWU45DLznH0Ar3oXuP9MAQAADUJlDwCwBZfJNn4gv1SHZA8AsAXzq94FbrIP3MgBAECDUNkDAGzBKYecJl6MY+ZafyPZAwBsgTY+AACwLCp7AIAtOGWuFe/0XSjNjmQPALAFO7fxSfYAAFtgIRwAAGBZVPYAAFswTK5nb/DoHQAALRttfAAA4FMbNmzQLbfcoqSkJDkcDq1YscLjuGEYmjJlihITExUREaH09HTt3LnT45zS0lINHz5c0dHRio2N1ciRI1VZWel1LCR7AIAtHF/i1szmjaqqKqWmpmr+/PknPT5nzhzl5uZq4cKF2rp1qyIjIzVgwABVV1e7zxk+fLg+//xzrVmzRqtWrdKGDRs0evRor387bXwAgC04Ta565+21AwcO1MCBA096zDAMzZ07V4888ogGDRokSVqyZIni4+O1YsUKDRs2TPn5+Vq9erU++ugj9e7dW5I0b9483XjjjXryySeVlJTU4Fio7AEA8EJ5ebnHVlNT4/U9CgsLVVxcrPT0dPe+mJgY9enTR1u2bJEkbdmyRbGxse5EL0np6ekKCgrS1q1bvfo+kj0AwBZ81cZPTk5WTEyMe8vJyfE6luLiYklSfHy8x/74+Hj3seLiYrVv397jeEhIiOLi4tznNBRtfACALbgUJJeJGvf4tUVFRYqOjnbvDwsLMx1bU6OyBwDAC9HR0R5bY5J9QkKCJKmkpMRjf0lJiftYQkKCDh065HH82LFjKi0tdZ/TUCR7AIAtOA2H6c1XUlJSlJCQoLVr17r3lZeXa+vWrUpLS5MkpaWlqaysTHl5ee5z1q1bJ5fLpT59+nj1fbTxAQC20JjH5/77em9UVlZq165d7s+FhYXasWOH4uLi1KFDB40bN06PPvqoOnfurJSUFGVnZyspKUmDBw+WJHXr1k033HCDRo0apYULF6qurk6ZmZkaNmyYVzPxJZI9AMAmDJOr3hleXrtt2zb169fP/TkrK0uSlJGRocWLF2vixImqqqrS6NGjVVZWpiuvvFKrV69WeHi4+5qlS5cqMzNT1113nYKCgjR06FDl5uZ6HTvJHgCAJtC3b18ZhnHK4w6HQzNmzNCMGTNOeU5cXJyWLVtmOhaSPQDAFpxyyGliMRsz1/obyR4AYAsuw/tx9/++PlAxGx8AAIujsoe+/rCN3n8hXgc/a62KQ6EatnCXuvU/7D7+xepYbVt2pg581lpHy0J076ovlNj9qPv4kbJgvTs3Sbs3RuvwgVBFxh1T1/5lunb8foVHu/zxk4DT+tWIPbr82kM6+5wq1dYEKf+TWC3KPV/7v4mUJLVPPKpF/9x40mtzJl6kTe9494wzWgaXyQl6Zq71N5I9VHckSAndjuqSX32v1+4778TjR4PUoXelLripVG9NPueE4xUlrVRR0koD/rhPZ3Y6qrL9YVr1SAdVlLTSr5/b0wy/APBOj14/6J9vJOurz2MUHGwoI3OnHn0uT/cOvVw11SH6riRcd15/jcc1NwzZpyG/+1rb3j/DT1HDLJcccpkYdzdzrb+1iGQ/f/58PfHEEyouLlZqaqrmzZunyy67zN9h2UbnvuXq3Lf8lMdTf1kqSfphX+hJj8d3qdawBT8m9biOtbpuwn69mZUi5zEpuEX8KQN+NCWzl8fnp6deqFfXvadO3cv1+fY4uVwO/fC951vR0vod0qY1Cao+yh9oBB6/9yRef/11ZWVlaerUqdq+fbtSU1M1YMCAE14RiMBSXRGssDZOEj0CQmTUMUlS5eFWJz3eqVu5zutaof9dcVZzhgUfa0lv0Gtufk/2Tz/9tEaNGqURI0aoe/fuWrhwoVq3bq1XXnnF36GhkapKg7V+XqJ6DfvO36EAP8vhMDR6wpf6/ONYfbM76qTn9B+0T3v3RCr//2KbNzj41PExezNboPJr5LW1tcrLy/NYzzcoKEjp6enu9Xx/qqam5oR1hNGyVFcEaenIzjqzc7X6jT3g73CAn3XfpHx1PK9Sj0++6KTHQ8OcumZgMVU9Appfk/13330np9N52vV8fyonJ8djDeHk5OTmChUNUFMZpP83orPCIp0atnC3gk/eEQVajHsfztdlV32ryaN76/tD4Sc954r0EoWFO7V2lXfvIkfL45LJ9ewDeIJeQPUkJk+erMOHD7u3oqIif4eE/6iuCNKSjM4KbmXojhd3qVVYAL99AjZg6N6H85XW75D++PveKjnQ+pRn9h+0X1vXn6nyspNPUEXgMP4zG7+xmxHAyd6v06fOOOMMBQcHn3Y9358KCwtr1LrBOL2aqiCVfvPjf9cfisJ08IsIRcQcU+xZdTpSFqzDB0JVUVJfqn+/p74CanNmnaLOPKbqiiD9NaOz6o4GaejTu1VTGayayvp7RcYdU1Bws/8k4LTun5SvawYWa+b4njp6JERt29VIkqoqQ1Rb8+Mf2MTkI7rwkh807cFL/BUqfKi5V71rSfya7ENDQ9WrVy+tXbvWvaSfy+XS2rVrlZmZ6c/QbOXAp621+Ddd3J/fnlU/PNJz6Hf65RPfqOCdWK2YeI77+N8ePFeS1PfBA+o37qAOft5a+3a0kST9uV8Pj3uP2/Cp2p5d28S/APDOTbfvkyQ9/tI2j/3PTL1A76z8cWz++kH79V1JuLZvades8QG+5vcHo7KyspSRkaHevXvrsssu09y5c1VVVaURI0b4OzTbSPlFpabvyTvl8Ytv+14X3/Z9o68HWpqbLunfoPOWPNtZS57t3MTRoLnwBj0/+vWvf61vv/1WU6ZMUXFxsXr27KnVq1efMGkPAAAzaOP7WWZmJm17AACaSItI9gAANDXejQ8AgMXZuY0fuLMNAABAg1DZAwBswc6VPckeAGALdk72tPEBALA4KnsAgC3YubIn2QMAbMGQucfnAnl5L5I9AMAW7FzZM2YPAIDFUdkDAGzBzpU9yR4AYAt2Tva08QEAsDgqewCALdi5sifZAwBswTAcMkwkbDPX+httfAAALI7KHgBgC6xnDwCAxdl5zJ42PgAAFkdlDwCwBTtP0CPZAwBswc5tfJI9AMAW7FzZM2YPAIDFUdkDAGzBMNnGD+TKnmQPALAFQ5JhmLs+UNHGBwDA4qjsAQC24JJDDt6gBwCAdTEbHwAAWBaVPQDAFlyGQw5eqgMAgHUZhsnZ+AE8HZ82PgAAFkdlDwCwBTtP0CPZAwBsgWQPAIDF2XmCHmP2AABYHJU9AMAW7Dwbn2QPALCF+mRvZszeh8E0M9r4AABYHJU9AMAWmI0PAIDFGTK3Jn0Ad/Fp4wMAYHVU9gAAW6CNDwCA1dm4j0+yBwDYg8nKXgFc2TNmDwCAxZHsAQC2cPwNemY2bzidTmVnZyslJUURERE677zzNHPmTBk/uZFhGJoyZYoSExMVERGh9PR07dy508e/nGQPALCJ4xP0zGzeePzxx7VgwQI9++yzys/P1+OPP645c+Zo3rx57nPmzJmj3NxcLVy4UFu3blVkZKQGDBig6upqn/52xuwBAGgCmzdv1qBBg3TTTTdJks455xy9+uqr+vDDDyXVV/Vz587VI488okGDBkmSlixZovj4eK1YsULDhg3zWSxU9gAAezAc5jdJ5eXlHltNTc1Jv+7yyy/X2rVr9dVXX0mSPvnkE23atEkDBw6UJBUWFqq4uFjp6enua2JiYtSnTx9t2bLFpz+dyh4AYAu+WvUuOTnZY//UqVM1bdq0E86fNGmSysvL1bVrVwUHB8vpdGrWrFkaPny4JKm4uFiSFB8f73FdfHy8+5ivkOwBAPBCUVGRoqOj3Z/DwsJOet4bb7yhpUuXatmyZbrgggu0Y8cOjRs3TklJScrIyGiucCWR7AEAduGjl+pER0d7JPtTeeihhzRp0iT32HuPHj30zTffKCcnRxkZGUpISJAklZSUKDEx0X1dSUmJevbsaSLQEzUo2b/11lsNvuGtt97a6GAAAGgqzf263CNHjigoyHNqXHBwsFwulyQpJSVFCQkJWrt2rTu5l5eXa+vWrbrvvvsaHefJNCjZDx48uEE3czgccjqdZuIBAMASbrnlFs2aNUsdOnTQBRdcoI8//lhPP/207r77bkn1OXPcuHF69NFH1blzZ6WkpCg7O1tJSUkNzrsN1aBkf/xfIQAABLRmfL/9vHnzlJ2drfvvv1+HDh1SUlKSfv/732vKlCnucyZOnKiqqiqNHj1aZWVluvLKK7V69WqFh4f7NBZTY/bV1dU+DwgAgKbQ3G38qKgozZ07V3Pnzj3lOQ6HQzNmzNCMGTMaHVdDeP2cvdPp1MyZM3XWWWepTZs22rNnjyQpOztbL7/8ss8DBADAJwwfbAHK62Q/a9YsLV68WHPmzFFoaKh7/4UXXqiXXnrJp8EBAADzvE72S5Ys0QsvvKDhw4crODjYvT81NVVffvmlT4MDAMB3HD7YApPXY/b79+9Xp06dTtjvcrlUV1fnk6AAAPA5Hz1nH4i8ruy7d++ujRs3nrD/73//uy6++GKfBAUAAHzH68p+ypQpysjI0P79++VyufSPf/xDBQUFWrJkiVatWtUUMQIAYB6VfcMNGjRIK1eu1DvvvKPIyEhNmTJF+fn5Wrlypa6//vqmiBEAAPN8tOpdIGrUc/ZXXXWV1qxZ4+tYAABAE2j0S3W2bdum/Px8SfXj+L169fJZUAAA+JqvlrgNRF4n+3379umOO+7Q+++/r9jYWElSWVmZLr/8cr322ms6++yzfR0jAADmMWbfcPfcc4/q6uqUn5+v0tJSlZaWKj8/Xy6XS/fcc09TxAgAAEzwurJfv369Nm/erC5durj3denSRfPmzdNVV13l0+AAAPAZs5Ps7DRBLzk5+aQvz3E6nUpKSvJJUAAA+JrDqN/MXB+ovG7jP/HEE3rggQe0bds2975t27Zp7NixevLJJ30aHAAAPmPjhXAaVNm3bdtWDseP7Yuqqir16dNHISH1lx87dkwhISG6++67NXjw4CYJFAAANE6Dkv3p1uIFACAgMGZ/ehkZGU0dBwAATcvGj941+qU6klRdXa3a2lqPfdHR0aYCAgAAvuX1BL2qqiplZmaqffv2ioyMVNu2bT02AABaJBtP0PM62U+cOFHr1q3TggULFBYWppdeeknTp09XUlKSlixZ0hQxAgBgno2Tvddt/JUrV2rJkiXq27evRowYoauuukqdOnVSx44dtXTpUg0fPrwp4gQAAI3kdWVfWlqqc889V1L9+Hxpaakk6corr9SGDRt8Gx0AAL5i4yVuvU725557rgoLCyVJXbt21RtvvCGpvuI/vjAOAAAtzfE36JnZApXXyX7EiBH65JNPJEmTJk3S/PnzFR4ervHjx+uhhx7yeYAAAMAcr8fsx48f7/7f6enp+vLLL5WXl6dOnTrpoosu8mlwAAD4DM/ZN17Hjh3VsWNHX8QCAACaQIOSfW5uboNv+OCDDzY6GAAAmopDJle981kkza9Byf6ZZ55p0M0cDgfJHgCAFqZByf747PuWavZFPRXiaOXvMIAm8faBt/0dAtBkyitcant+M30ZC+EAAGBxNp6g5/WjdwAAILBQ2QMA7MHGlT3JHgBgC2bfgmerN+gBAIDA0qhkv3HjRt15551KS0vT/v37JUl//etftWnTJp8GBwCAz9h4iVuvk/2bb76pAQMGKCIiQh9//LFqamokSYcPH9bs2bN9HiAAAD5Bsm+4Rx99VAsXLtSLL76oVq1+fLb9iiuu0Pbt230aHAAAMM/rCXoFBQW6+uqrT9gfExOjsrIyX8QEAIDPMUHPCwkJCdq1a9cJ+zdt2qRzzz3XJ0EBAOBzx9+gZ2YLUF4n+1GjRmns2LHaunWrHA6HDhw4oKVLl2rChAm67777miJGAADMs/GYvddt/EmTJsnlcum6667TkSNHdPXVVyssLEwTJkzQAw880BQxAgAAE7xO9g6HQ3/605/00EMPadeuXaqsrFT37t3Vpk2bpogPAACfsPOYfaPfoBcaGqru3bv7MhYAAJoOr8ttuH79+snhOPUkhXXr1pkKCAAA+JbXyb5nz54en+vq6rRjxw599tlnysjI8FVcAAD4lsk2vq0q+2eeeeak+6dNm6bKykrTAQEA0CRs3Mb32UI4d955p1555RVf3Q4AAPiIz5a43bJli8LDw311OwAAfMvGlb3XyX7IkCEenw3D0MGDB7Vt2zZlZ2f7LDAAAHyJR++8EBMT4/E5KChIXbp00YwZM9S/f3+fBQYAAHzDq2TvdDo1YsQI9ejRQ23btm2qmAAAgA95NUEvODhY/fv3Z3U7AEDgsfG78b2ejX/hhRdqz549TRELAABN5viYvZktUHmd7B999FFNmDBBq1at0sGDB1VeXu6xAQCAlqXBY/YzZszQH/7wB914442SpFtvvdXjtbmGYcjhcMjpdPo+SgAAfCGAq3MzGpzsp0+frnvvvVfvvvtuU8YDAEDT4Dn7n2cY9b/ymmuuabJgAACA73n16N3pVrsDAKAl46U6DXT++ef/bMIvLS01FRAAAE2CNn7DTJ8+/YQ36AEAgJbNq2Q/bNgwtW/fvqliAQCgydDGbwDG6wEAAc3GbfwGv1Tn+Gx8AAAQWBqc7F0uFy18AEDg8sO78ffv368777xT7dq1U0REhHr06KFt27b9GJJhaMqUKUpMTFRERITS09O1c+dOEz/y5Lx+XS4AAIGoud+N/8MPP+iKK65Qq1at9O9//1tffPGFnnrqKY9VY+fMmaPc3FwtXLhQW7duVWRkpAYMGKDq6mqf/nav17MHACAg+WjM/r/XgQkLC1NYWNgJpz/++ONKTk7WokWL3PtSUlJ+vJ1haO7cuXrkkUc0aNAgSdKSJUsUHx+vFStWaNiwYSaC9URlDwCAF5KTkxUTE+PecnJyTnreW2+9pd69e+tXv/qV2rdvr4svvlgvvvii+3hhYaGKi4uVnp7u3hcTE6M+ffpoy5YtPo2Zyh4AYA8+quyLiooUHR3t3n2yql6S9uzZowULFigrK0t//OMf9dFHH+nBBx9UaGioMjIyVFxcLEmKj4/3uC4+Pt59zFdI9gAAW/DVc/bR0dEeyf5UXC6XevfurdmzZ0uSLr74Yn322WdauHChMjIyGh9II9DGBwCgCSQmJqp79+4e+7p166a9e/dKkhISEiRJJSUlHueUlJS4j/kKyR4AYA/N/OjdFVdcoYKCAo99X331lTp27CipfrJeQkKC1q5d6z5eXl6urVu3Ki0tzeufdzq08QEAttDcr8sdP368Lr/8cs2ePVu33367PvzwQ73wwgt64YUX6u/ncGjcuHF69NFH1blzZ6WkpCg7O1tJSUkaPHhw4wM9CZI9AABN4NJLL9Xy5cs1efJkzZgxQykpKZo7d66GDx/uPmfixImqqqrS6NGjVVZWpiuvvFKrV69WeHi4T2Mh2QMA7MEP78a/+eabdfPNN5/yuMPh0IwZMzRjxgwTgf08kj0AwB5YCAcAAFgVlT0AwBYc/9nMXB+oSPYAAHuwcRufZA8AsIXmfvSuJWHMHgAAi6OyBwDYA218AABsIIATthm08QEAsDgqewCALdh5gh7JHgBgDzYes6eNDwCAxVHZAwBsgTY+AABWRxsfAABYFZU9AMAWaOMDAGB1Nm7jk+wBAPZg42TPmD0AABZHZQ8AsAXG7AEAsDra+AAAwKqo7AEAtuAwDDmMxpfnZq71N5I9AMAeaOMDAACrorIHANgCs/EBALA62vgAAMCqqOwBALZAGx8AAKuzcRufZA8AsAU7V/aM2QMAYHFU9gAAe6CNDwCA9QVyK94M2vgAAFgclT0AwB4Mo34zc32AItkDAGyB2fgAAMCyqOwBAPbAbHwAAKzN4arfzFwfqGjjAwBgcVT2OMGvM0t0xY2HldypRrXVQfpiW2u9PCtR+3aHu89pFebS6KkH1PfWMrUKM5T3XpTmTT5LZd+18mPkwMl9+kGk/vZce+38tLVKS1pp6suFunzgYUnSsTpp8eOJ+mhdtA5+E6rIaJcuvqpCI/94QO0SjrnvsezP8frwnWjt+TxCIaGG/vHlp/76OWgsG7fxqexxgovSqrRy8Rkad3NnTR52roJDDM1+dY/CIpzuc+6ddkC/uL5cj/6+oyYMOU9x8XWa8vLX/gsaOI3qI0E694Kjypy974RjNUeDtOvT1vrNuBLNf/srTXmpUPt2h2nqXed6nHes1qGrbynTTRnfNVfY8LHjs/HNbIHKr5X9hg0b9MQTTygvL08HDx7U8uXLNXjwYH+GBEl/Gu75l9xT4zrojc8+V+eLjuqzrW3UOsqpAXeU6rExHfTJ+1GSpKezkvXShgJ1vaRKX26P9EfYwCldem2FLr224qTHIqNdeuz13R77xszapwdv7KJD+1qp/dl1kqTfPVQsSfrf1+OaNlg0HRs/Z+/Xyr6qqkqpqamaP3++P8PAz4iMrq/oK8qCJUmdLzqiVqGGPt4Y5T6naFe4Sva1UrdeR/wSI+BLVeXBcjgMRcY4f/5kIAD4tbIfOHCgBg4c2ODza2pqVFNT4/5cXl7eFGHhJxwOQ/dO36/PPmytbwoiJElx7Y+ptsahqvJgj3PLvg1RXPs6f4QJ+ExttUMvz0pS38E/KDIqgKdf4wS8VCdA5OTkKCYmxr0lJyf7OyTLy5y9Xx27Vivnvo7+DgVocsfqpFm/P0cypAceO3F8HwHO8MEWoAIq2U+ePFmHDx92b0VFRf4OydLGzNqnPteXa+Jt5+m7g6Hu/aWHQhQaZrjb+8fFnnlMpYeYjY/AdDzRl+wPVc5ru6nqYSkB9ehdWFiYwsLC/B2GDRgaM2u/Lr/hsB66rZNKijz/m+/8v9aqq3Xo4isrtOlfsZKks8+rVvzZdcrPa+2HeAFzjif6/YVhmvP3XYqOY6zeiuzcxg+oZI/mkTl7v/r98gdNG5Gio5VBantm/Th8VUWwaquDdKQiWG+/GqfR0w6ooixEVRVBGjNrv77Y1pqZ+GiRjlYF6UDhj/9oLS4K1e7PIhQVe0xx8XWaOSpFuz6N0Iwle+RyOlR6qP6vxqhYp1qF1v8Nf2hfK1WUhejQ/lZyOaXdn9XPYUlKqVFEJF2AgGDj2fgke5zglru+lyQ9+Q/Px5GeHJesNW/UP3a0cFqSXIaU/eLXahVmaNt7UXp28lnNHivQEF990loTb+vk/vz8tPo/q9ffXqo7/1CsD/43RpJ0//VdPa6b8/ddSr28UpK05MlE959/Sbq/f5cTzgFaKr8m+8rKSu3atcv9ubCwUDt27FBcXJw6dOjgx8jsbUBS6s+eU1cTpPl/PFvz/3h2M0QEmJN6eaXePrDjlMdPd+y4CXP3asLcvb4LCs2ONr6fbNu2Tf369XN/zsrKkiRlZGRo8eLFfooKAGBJNn5drl+Tfd++fWUE8BgIAACBgDF7AIAt0MYHAMDqXEb9Zub6AEWyBwDYg43H7APqDXoAAMB7VPYAAFtwyOSYvc8iaX4kewCAPdj4DXq08QEAsDgqewCALdj50TsqewCAPfhxPfvHHntMDodD48aNc++rrq7WmDFj1K5dO7Vp00ZDhw5VSUlJ47/kNEj2AAA0oY8++kjPP/+8LrroIo/948eP18qVK/W3v/1N69ev14EDBzRkyJAmiYFkDwCwBYdhmN68VVlZqeHDh+vFF19U27Zt3fsPHz6sl19+WU8//bSuvfZa9erVS4sWLdLmzZv1wQcf+PJnSyLZAwDswuWDTVJ5ebnHVlNTc8qvHDNmjG666Salp6d77M/Ly1NdXZ3H/q5du6pDhw7asmWLT37uT5HsAQDwQnJysmJiYtxbTk7OSc977bXXtH379pMeLy4uVmhoqGJjYz32x8fHq7i42OcxMxsfAGALjW3F//R6SSoqKlJ0dLR7f1hY2AnnFhUVaezYsVqzZo3Cw8Mb/Z2+QmUPALAHH83Gj46O9thOluzz8vJ06NAhXXLJJQoJCVFISIjWr1+v3NxchYSEKD4+XrW1tSorK/O4rqSkRAkJCT7/6VT2AAB7aMY36F133XX69NNPPfaNGDFCXbt21cMPP6zk5GS1atVKa9eu1dChQyVJBQUF2rt3r9LS0hof4ymQ7AEA8LGoqChdeOGFHvsiIyPVrl079/6RI0cqKytLcXFxio6O1gMPPKC0tDT94he/8Hk8JHsAgC20tDfoPfPMMwoKCtLQoUNVU1OjAQMG6LnnnvPtl/wHyR4AYA9+Xgjnvffe8/gcHh6u+fPna/78+abu2xBM0AMAwOKo7AEAtuBw1W9mrg9UJHsAgD2wnj0AALAqKnsAgD2YXKbW1LV+RrIHANiCr16XG4ho4wMAYHFU9gAAe7DxBD2SPQDAHgy516Rv9PUBimQPALAFxuwBAIBlUdkDAOzBkMkxe59F0uxI9gAAe7DxBD3a+AAAWByVPQDAHlySHCavD1AkewCALTAbHwAAWBaVPQDAHmw8QY9kDwCwBxsne9r4AABYHJU9AMAebFzZk+wBAPbAo3cAAFgbj94BAADLorIHANgDY/YAAFicy5AcJhK2K3CTPW18AAAsjsoeAGAPtPEBALA6k8legZvsaeMDAGBxVPYAAHugjQ8AgMW5DJlqxTMbHwAAtFRU9gAAezBc9ZuZ6wMUyR4AYA+M2QMAYHGM2QMAAKuisgcA2ANtfAAALM6QyWTvs0iaHW18AAAsjsoeAGAPtPEBALA4l0uSiWflXYH7nD1tfAAALI7KHgBgD7TxAQCwOBsne9r4AABYHJU9AMAebPy6XJI9AMAWDMMlw8TKdWau9TeSPQDAHgzDXHXOmD0AAGipqOwBAPZgmByzD+DKnmQPALAHl0tymBh3D+Axe9r4AABYHJU9AMAeaOMDAGBthsslw0QbP5AfvaONDwCAxVHZAwDsgTY+AAAW5zIkhz2TPW18AAAsjsoeAGAPhiHJzHP2gVvZk+wBALZguAwZJtr4BskeAIAWznDJXGXPo3cAAOAncnJydOmllyoqKkrt27fX4MGDVVBQ4HFOdXW1xowZo3bt2qlNmzYaOnSoSkpKfB4LyR4AYAuGyzC9eWP9+vUaM2aMPvjgA61Zs0Z1dXXq37+/qqqq3OeMHz9eK1eu1N/+9jetX79eBw4c0JAhQ3z902njAwBsopnb+KtXr/b4vHjxYrVv3155eXm6+uqrdfjwYb388statmyZrr32WknSokWL1K1bN33wwQf6xS9+0fhY/0tAJ/vjkyWOqc7UexKAlqy8InDHCYGfU15Z/+e7OSa/mc0Vx1QnSSovL/fYHxYWprCwsJ+9/vDhw5KkuLg4SVJeXp7q6uqUnp7uPqdr167q0KGDtmzZQrI/rqKiQpK0Sf/ycyRA02l7vr8jAJpeRUWFYmJimuTeoaGhSkhI0KZi87miTZs2Sk5O9tg3depUTZs27bTXuVwujRs3TldccYUuvPBCSVJxcbFCQ0MVGxvrcW58fLyKi4tNx/pTAZ3sk5KSVFRUpKioKDkcDn+HYwvl5eVKTk5WUVGRoqOj/R0O4FP8+W5+hmGooqJCSUlJTfYd4eHhKiwsVG1trel7GYZxQr5pSFU/ZswYffbZZ9q0aZPpGBojoJN9UFCQzj77bH+HYUvR0dH8ZQjL4s9382qqiv6nwsPDFR4e3uTfczKZmZlatWqVNmzY4JGzEhISVFtbq7KyMo/qvqSkRAkJCT6Ngdn4AAA0AcMwlJmZqeXLl2vdunVKSUnxON6rVy+1atVKa9eude8rKCjQ3r17lZaW5tNYArqyBwCgpRozZoyWLVum//mf/1FUVJR7HD4mJkYRERGKiYnRyJEjlZWVpbi4OEVHR+uBBx5QWlqaTyfnSSR7eCksLExTp05t0BgVEGj48w1fWrBggSSpb9++HvsXLVqku+66S5L0zDPPKCgoSEOHDlVNTY0GDBig5557zuexOIxAftkvAAD4WYzZAwBgcSR7AAAsjmQPAIDFkewBALA4kj0abP78+TrnnHMUHh6uPn366MMPP/R3SIBPbNiwQbfccouSkpLkcDi0YsUKf4cE+BTJHg3y+uuvKysrS1OnTtX27duVmpqqAQMG6NChQ/4ODTCtqqpKqampmj9/vr9DAZoEj96hQfr06aNLL71Uzz77rKT6RR2Sk5P1wAMPaNKkSX6ODvAdh8Oh5cuXa/Dgwf4OBfAZKnv8rNraWuXl5XkswxgUFKT09HRt2bLFj5EBABqCZI+f9d1338npdCo+Pt5jf1MswwgA8D2SPQAAFkeyx88644wzFBwcrJKSEo/9TbEMIwDA90j2+FmhoaHq1auXxzKMLpdLa9eu9fkyjAAA32PVOzRIVlaWMjIy1Lt3b1122WWaO3euqqqqNGLECH+HBphWWVmpXbt2uT8XFhZqx44diouLU4cOHfwYGeAbPHqHBnv22Wf1xBNPqLi4WD179lRubq769Onj77AA09577z3169fvhP0ZGRlavHhx8wcE+BjJHgAAi2PMHgAAiyPZAwBgcSR7AAAsjmQPAIDFkewBALA4kj0AABZHsgcAwOJI9gAAWBzJHjDprrvu0uDBg92f+/btq3HjxjV7HO+9954cDofKyspOeY7D4dCKFSsafM9p06apZ8+epuL6+uuv5XA4tGPHDlP3AdB4JHtY0l133SWHwyGHw6HQ0FB16tRJM2bM0LFjx5r8u//xj39o5syZDTq3IQkaAMxiIRxY1g033KBFixappqZG//rXvzRmzBi1atVKkydPPuHc2tpahYaG+uR74+LifHIfAPAVKntYVlhYmBISEtSxY0fdd999Sk9P11tvvSXpx9b7rFmzlJSUpC5dukiSioqKdPvttys2NlZxcXEaNGiQvv76a/c9nU6nsrKyFBsbq3bt2mnixIn67+Ul/ruNX1NTo4cffljJyckKCwtTp06d9PLLL+vrr792L77Stm1bORwO3XXXXZLqlxDOyclRSkqKIiIilJqaqr///e8e3/Ovf/1L559/viIiItSvXz+POBvq4Ycf1vnnn6/WrVvr3HPPVXZ2turq6k447/nnn1dycrJat26t22+/XYcPH/Y4/tJLL6lbt24KDw9X165d9dxzz3kdC4CmQ7KHbURERKi2ttb9ee3atSooKNCaNWu0atUq1dXVacCAAYqKitLGjRv1/vvvq02bNrrhhhvc1z311FNavHixXnnlFW3atEmlpaVavnz5ab/3d7/7nV599VXl5uYqPz9fzz//vNq0aaPk5GS9+eabkqSCggIdPHhQf/7znyVJOTk5WrJkiRYuXKjPP/9c48eP15133qn169dLqv9HyZAhQ3TLLbdox44duueeezRp0iSv/5tERUVp8eLF+uKLL/TnP/9ZL774op555hmPc3bt2qU33nhDK1eu1OrVq/Xxxx/r/vvvdx9funSppkyZolmzZik/P1+zZ89Wdna2/vKXv3gdD4AmYgAWlJGRYQwaNMgwDMNwuVzGmjVrjLCwMGPChAnu4/Hx8UZNTY37mr/+9a9Gly5dDJfL5d5XU1NjREREGG+//bZhGIaRmJhozJkzx328rq7OOPvss93fZRiGcc011xhjx441DMMwCgoKDEnGmjVrThrnu+++a0gyfvjhB/e+6upqo3Xr1sbmzZs9zh05cqRxxx13GIZhGJMnTza6d+/ucfzhhx8+4V7/TZKxfPnyUx5/4oknjF69erk/T5061QgODjb27dvn3vfvf//bCAoKMg4ePGgYhmGcd955xrJlyzzuM3PmTCMtLc0wDMMoLCw0JBkff/zxKb8XQNNizB6WtWrVKrVp00Z1dXVyuVz6zW9+o2nTprmP9+jRw2Oc/pNPPtGuXbsUFRXlcZ/q6mrt3r1bhw8f1sGDB9WnTx/3sZCQEPXu3fuEVv5xO3bsUHBwsK655poGx71r1y4dOXJE119/vcf+2tpaXXzxxZKk/Px8jzgkKS0trcHfcdzrr7+u3Nxc7d69W5WVlTp27Jiio6M9zunQoYPOOussj+9xuVwqKChQVFSUdu/erZEjR2rUqFHuc44dO6aYmBiv4wHQNEj2sKx+/fppwYIFCg0NVVJSkkJCPP+4R0ZGenyurKxUr169tHTp0hPudeaZZzYqhoiICK+vqayslCT985//9EiyUv08BF/ZsmWLhg8frunTp2vAgAGKiYnRa6+9pqeeesrrWF988cUT/vERHBzss1gBmEOyh2VFRkaqU6dODT7/kksu0euvv6727dufUN0el5iYqK1bt+rqq6+WVF/B5uXl6ZJLLjnp+T169JDL5dL69euVnp5+wvHjnQWn0+ne1717d4WFhWnv3r2n7Ah069bNPdnwuA8++ODnf+RPbN68WR07dtSf/vQn975vvvnmhPP27t2rAwcOKCkpyf09QUFB6tKli+Lj45WUlKQ9e/Zo+PDhXn0/gObDBD3gP4YPH64zzjhDgwYN0saNG1VYWKj33ntPDz74oPbt2ydJGjt2rB577DGtWLFCX375pe6///7TPiN/zjnnKCMjQ3fffbdWrFjhvucbb7whSerYsaMcDodWrVqlb7/9VpWVlYqKitKECRM0fvx4/eUvf9Hu3bu1fft2zZs3zz3p7d5779XOnTv10EMPqaCgQMuWLdPixYu9+r2dO3fW3r179dprr2n37t3Kzc096WTD8PBwZWRk6JNPPtHGjRv14IMP6vbbb1dCQoIkafr06crJyVFubq6++uorffrpp1q0aJGefvppr+IB0HRI9sB/tG7dWhs2bFCHDh00ZMgQdevWTSNHjlR1dbW70v/DH/6g3/72t8rIyFBaWpqioqL0y1/+8rT3XbBggW677Tbdf//96tq1q0aNGqWqqipJ0llnnaXp06dr0qRJio+PV2ZmpiRp5syZys7OVk5Ojrp166YbbrhB//znP5WSkiKpfhz9zTff1IoVK5SamqqFCxdq9uzZXv3eW2+9VePHj1dmZqZ69uypzZs3Kzs7+4TzOnXqpCFDhujGG29U//79ddFFF3k8WnfPPffopZde0qJFi9SjRw9dc801Wrx4sTtWAP7nME41swgAAFgClT0AABZHsgcAwOJI9gAAWBzJHgAAiyPZAwBgcSR7AAAsjmQPAIDFkewBALA4kj0AABZHsgcAwOJI9gAAWNz/B7GOHIyG7sVHAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fold 1 AUC: 0.8541\n",
      "Fold 2 AUC: 0.8069\n",
      "Fold 3 AUC: 0.8104\n",
      "Fold 4 AUC: 0.8614\n",
      "Fold 5 AUC: 0.9180\n",
      "Fold 6 AUC: 0.9990\n",
      "Fold 7 AUC: 0.9867\n",
      "Fold 8 AUC: 0.9820\n",
      "Fold 9 AUC: 0.9659\n",
      "Fold 10 AUC: 0.9855\n",
      "Mean AUC: 0.9170\n"
     ]
    }
   ],
   "source": [
    "# Confusion Matrix for Gradient Boosting\n",
    "gb_confusion_matrix = confusion_matrix(Classification_y_test, classification_model_gb_predict)\n",
    "tn, fp, fn, tp = gb_confusion_matrix.ravel()\n",
    "recall = tp /(tp + fn)\n",
    "precision = tp / (tp + fp)\n",
    "f_measure = ((2* recall)*precision)/(recall + precision)\n",
    "print(f\"Recall: {recall}\")\n",
    "print(f\"Precision: {precision}\")\n",
    "print(f\"F-measure: {f_measure}\")\n",
    "print(\"Confusion Matrix for Gradient Boosting:\")\n",
    "ConfusionMatrixDisplay(gb_confusion_matrix).plot()\n",
    "\n",
    "#ROC Cure and AUC and 10 cross validation\n",
    "cv = StratifiedKFold(n_splits=10)\n",
    "\n",
    "roc_curves = []\n",
    "auc_scores = []\n",
    "\n",
    "\n",
    "for train_idx, test_idx in cv.split(Classification_X,Classification_y):\n",
    "    X_train, X_test = Classification_X.iloc[train_idx], Classification_X.iloc[test_idx]\n",
    "    y_train, y_test = Classification_y.iloc[train_idx], Classification_y.iloc[test_idx]\n",
    "    classification_model_gb.fit(X_train, y_train)\n",
    "    y_score = classification_model_gb.predict_proba(X_test)[:, 1]\n",
    "    fpr, tpr, _ = roc_curve(y_test, y_score)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    roc_curves.append((fpr, tpr, roc_auc))\n",
    "    auc_scores.append(roc_auc)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "for fpr, tpr, roc_auc in roc_curves:\n",
    "    plt.plot(fpr, tpr, lw=1, alpha=0.7)\n",
    "mean_fpr = np.linspace(0, 1, 100)\n",
    "mean_tpr = np.mean([np.interp(mean_fpr, fpr, tpr) for fpr, tpr, _ in roc_curves], axis=0)\n",
    "plt.plot(mean_fpr, mean_tpr, color='b', label='Mean ROC')\n",
    "plt.plot([0, 1], [0, 1], color='gray', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()\n",
    "for fold, auc_score in enumerate(auc_scores, start=1):\n",
    "    print(f\"Fold {fold} AUC: {auc_score:.4f}\")\n",
    "mean_auc = np.mean(auc_scores)\n",
    "print(f\"Mean AUC: {mean_auc:.4f}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "steps: \n",
    "    I have loaded the data and then split the data as training set(80%) for training the model and test set(20%) for testing the training model performance.\n",
    "\n",
    "    Then I used RandomizedSearchCV to figure the best parameter for different models. \n",
    "\n",
    "    I have used three model: logistic regressionn (regression algorithm), gradient boosting and random forrest (the last two are ensembles)\n",
    "\n",
    "    I have print out the recall, percision, and F-measure confutsion matrix, ROC curve and AUC result.\n",
    "\n",
    "    I also done 10-cross validation to make the result more reliable. \n",
    "\n",
    "Result:\n",
    "    ROC is the probability curve and AUC represents the degree or measure of separability. Higher the AUCm the better the model predict class good as good and class bad as bad. \n",
    "\n",
    "    From the ROC curve we can see that the ensemble algorithm performs relatively better performance than regression algorithm, and graient boosting done the best among others. \n",
    "\n",
    "    For the mean value of AUC we could see that ensemnle algorithm (gradient boosting and random forrest) have around 90% chance that ditinguish between postive and negative class. Well the regression algorithm only performs 77% chance to get them distinguished.\n",
    "    \n",
    "    The reason may due to algorithm differecen as regression algorithm could handle errors but the result may skew. While ensembels algorithm is more robust in this case.\n",
    "\n",
    "Interesting finding:\n",
    "\n",
    "    For imbalanced classification problem target, we should do resampling, otherwise, even random guess can get a relatively good interpretation. And other parameter tuning does not enhance much of the prediction performance. \n",
    "    \n",
    "\n",
    "    \n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv-msa",
   "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
}