Genetic Algorithm for Titanic Data Set.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Genetic Algorithm for Titanic Data Set\n",
    "\n",
    "Train a feedforward neural network (with sigmoidal node functions and a hidden layer with twice as many nodes as the input dimensionality) for a 2-class classification task using GA/ES/PSO;  compare with backpropagation.\n",
    "\n",
    "https://www.kaggle.com/c/titanic/data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import pandas as pd \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv('./titanic_dataset/train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>714.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.383838</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>32.204208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>257.353842</td>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>14.526497</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>49.693429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>223.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>20.125000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.910400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.454200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>668.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>31.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>512.329200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId    Survived      Pclass         Age       SibSp  \\\n",
       "count   891.000000  891.000000  891.000000  714.000000  891.000000   \n",
       "mean    446.000000    0.383838    2.308642   29.699118    0.523008   \n",
       "std     257.353842    0.486592    0.836071   14.526497    1.102743   \n",
       "min       1.000000    0.000000    1.000000    0.420000    0.000000   \n",
       "25%     223.500000    0.000000    2.000000   20.125000    0.000000   \n",
       "50%     446.000000    0.000000    3.000000   28.000000    0.000000   \n",
       "75%     668.500000    1.000000    3.000000   38.000000    1.000000   \n",
       "max     891.000000    1.000000    3.000000   80.000000    8.000000   \n",
       "\n",
       "            Parch        Fare  \n",
       "count  891.000000  891.000000  \n",
       "mean     0.381594   32.204208  \n",
       "std      0.806057   49.693429  \n",
       "min      0.000000    0.000000  \n",
       "25%      0.000000    7.910400  \n",
       "50%      0.000000   14.454200  \n",
       "75%      0.000000   31.000000  \n",
       "max      6.000000  512.329200  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.0      1   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0      1   \n",
       "2                             Heikkinen, Miss. Laina  female  26.0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0      1   \n",
       "4                           Allen, Mr. William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked  \n",
       "0      0         A/5 21171   7.2500   NaN        S  \n",
       "1      0          PC 17599  71.2833   C85        C  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      0            113803  53.1000  C123        S  \n",
       "4      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dropping the columns that dont have co-relation with survival"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.drop(['PassengerId', 'Name', 'Ticket', 'Cabin'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass     Sex   Age  SibSp  Parch     Fare Embarked\n",
       "0         0       3    male  22.0      1      0   7.2500        S\n",
       "1         1       1  female  38.0      1      0  71.2833        C\n",
       "2         1       3  female  26.0      0      0   7.9250        S\n",
       "3         1       1  female  35.0      1      0  53.1000        S\n",
       "4         0       3    male  35.0      0      0   8.0500        S"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(891, 8)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dropping rows that do not have age values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.dropna(subset=['Age'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(714, 8)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing Value Embarked count:  2\n",
      "Missing Value Pclass count:  0\n",
      "Missing Value Sex count:  0\n",
      "Missing Value SibSp count:  0\n",
      "Missing Value Parch count:  0\n",
      "Missing Value Fare count:  0\n"
     ]
    }
   ],
   "source": [
    "print(\"Missing Value Embarked count: \",train.Embarked.isnull().sum())\n",
    "print(\"Missing Value Pclass count: \",train.Pclass.isnull().sum())\n",
    "print(\"Missing Value Sex count: \",train.Sex.isnull().sum())\n",
    "print(\"Missing Value SibSp count: \",train.SibSp.isnull().sum())\n",
    "print(\"Missing Value Parch count: \",train.Parch.isnull().sum())\n",
    "print(\"Missing Value Fare count: \",train.Fare.isnull().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['Embarked'].fillna('S',inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Converting non numeric Embarked values and Sex values to numeric type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convertEmbarkedValues(val):\n",
    "    if val == 'S':\n",
    "        return 1\n",
    "    elif val == 'C':\n",
    "        return 2\n",
    "    else:\n",
    "        return 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convertSexValues(val):\n",
    "    if val == 'male':\n",
    "        return 0\n",
    "    elif val == 'female':\n",
    "        return 1\n",
    "    else:\n",
    "        return 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['Embarked_Val'] = train.apply(lambda row: convertEmbarkedValues(row['Embarked']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['Sex_Val'] = train.apply(lambda row: convertSexValues(row['Sex']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.drop(['Embarked', 'Sex'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked_Val</th>\n",
       "      <th>Sex_Val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass   Age  SibSp  Parch     Fare  Embarked_Val  Sex_Val\n",
       "0         0       3  22.0      1      0   7.2500             1        0\n",
       "1         1       1  38.0      1      0  71.2833             2        1\n",
       "2         1       3  26.0      0      0   7.9250             1        1\n",
       "3         1       1  35.0      1      0  53.1000             1        1\n",
       "4         0       3  35.0      0      0   8.0500             1        0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Neural network using back-propagation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train, x_test, y_train, y_test = train_test_split(train.drop('Survived',axis=1), train['Survived'], test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "layer_1 (Dense)              (None, 32)                256       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "layer_2 (Dense)              (None, 1)                 33        \n",
      "=================================================================\n",
      "Total params: 289\n",
      "Trainable params: 289\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(32, input_dim=7, activation='sigmoid', name='layer_1'))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(1, activation='sigmoid', name='layer_2'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7, 32)\n",
      "(32,)\n",
      "(32, 1)\n",
      "(1,)\n"
     ]
    }
   ],
   "source": [
    "testLayer1 = model.get_layer('layer_1').get_weights()\n",
    "print(testLayer1[0].shape)\n",
    "print(testLayer1[1].shape)\n",
    "testLayer2= model.get_layer('layer_2').get_weights()\n",
    "print(testLayer2[0].shape)\n",
    "print(testLayer2[1].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testLayer1[1] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='binary_crossentropy', optimizer='rmsprop', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 513 samples, validate on 58 samples\n",
      "Epoch 1/50\n",
      "513/513 [==============================] - 2s 4ms/step - loss: 0.6386 - acc: 0.6511 - val_loss: 0.5957 - val_acc: 0.7069\n",
      "Epoch 2/50\n",
      "513/513 [==============================] - 0s 723us/step - loss: 0.6145 - acc: 0.6862 - val_loss: 0.5887 - val_acc: 0.7069\n",
      "Epoch 3/50\n",
      "513/513 [==============================] - 0s 569us/step - loss: 0.6097 - acc: 0.6998 - val_loss: 0.5896 - val_acc: 0.7069\n",
      "Epoch 4/50\n",
      "513/513 [==============================] - 0s 715us/step - loss: 0.6001 - acc: 0.7018 - val_loss: 0.5935 - val_acc: 0.7069\n",
      "Epoch 5/50\n",
      "513/513 [==============================] - 0s 546us/step - loss: 0.5883 - acc: 0.6940 - val_loss: 0.5782 - val_acc: 0.7241\n",
      "Epoch 6/50\n",
      "513/513 [==============================] - 0s 610us/step - loss: 0.5918 - acc: 0.7037 - val_loss: 0.5733 - val_acc: 0.7241\n",
      "Epoch 7/50\n",
      "513/513 [==============================] - 0s 601us/step - loss: 0.5877 - acc: 0.7018 - val_loss: 0.5755 - val_acc: 0.7241\n",
      "Epoch 8/50\n",
      "513/513 [==============================] - 0s 481us/step - loss: 0.5725 - acc: 0.7076 - val_loss: 0.5760 - val_acc: 0.7241\n",
      "Epoch 9/50\n",
      "513/513 [==============================] - 0s 557us/step - loss: 0.5756 - acc: 0.7173 - val_loss: 0.5715 - val_acc: 0.7414\n",
      "Epoch 10/50\n",
      "513/513 [==============================] - 0s 575us/step - loss: 0.5790 - acc: 0.7096 - val_loss: 0.5700 - val_acc: 0.7414\n",
      "Epoch 11/50\n",
      "513/513 [==============================] - 0s 590us/step - loss: 0.5711 - acc: 0.7057 - val_loss: 0.5566 - val_acc: 0.7414\n",
      "Epoch 12/50\n",
      "513/513 [==============================] - 0s 501us/step - loss: 0.5709 - acc: 0.7096 - val_loss: 0.5612 - val_acc: 0.7414\n",
      "Epoch 13/50\n",
      "513/513 [==============================] - 0s 533us/step - loss: 0.5662 - acc: 0.7212 - val_loss: 0.5664 - val_acc: 0.7414\n",
      "Epoch 14/50\n",
      "513/513 [==============================] - 0s 441us/step - loss: 0.5661 - acc: 0.7037 - val_loss: 0.5618 - val_acc: 0.7414\n",
      "Epoch 15/50\n",
      "513/513 [==============================] - 0s 440us/step - loss: 0.5520 - acc: 0.7115 - val_loss: 0.5503 - val_acc: 0.7414\n",
      "Epoch 16/50\n",
      "513/513 [==============================] - 0s 555us/step - loss: 0.5514 - acc: 0.7173 - val_loss: 0.5419 - val_acc: 0.7414\n",
      "Epoch 17/50\n",
      "513/513 [==============================] - 0s 557us/step - loss: 0.5555 - acc: 0.6998 - val_loss: 0.5412 - val_acc: 0.7414\n",
      "Epoch 18/50\n",
      "513/513 [==============================] - 0s 570us/step - loss: 0.5488 - acc: 0.7310 - val_loss: 0.5382 - val_acc: 0.7414\n",
      "Epoch 19/50\n",
      "513/513 [==============================] - 0s 411us/step - loss: 0.5535 - acc: 0.7173 - val_loss: 0.5393 - val_acc: 0.7414\n",
      "Epoch 20/50\n",
      "513/513 [==============================] - 0s 378us/step - loss: 0.5446 - acc: 0.7329 - val_loss: 0.5424 - val_acc: 0.7586\n",
      "Epoch 21/50\n",
      "513/513 [==============================] - 0s 413us/step - loss: 0.5260 - acc: 0.7368 - val_loss: 0.5412 - val_acc: 0.7759\n",
      "Epoch 22/50\n",
      "513/513 [==============================] - 0s 502us/step - loss: 0.5338 - acc: 0.7446 - val_loss: 0.5336 - val_acc: 0.7759\n",
      "Epoch 23/50\n",
      "513/513 [==============================] - 0s 410us/step - loss: 0.5378 - acc: 0.7271 - val_loss: 0.5378 - val_acc: 0.7759\n",
      "Epoch 24/50\n",
      "513/513 [==============================] - 0s 384us/step - loss: 0.5227 - acc: 0.7583 - val_loss: 0.5278 - val_acc: 0.7759\n",
      "Epoch 25/50\n",
      "513/513 [==============================] - 0s 385us/step - loss: 0.5256 - acc: 0.7466 - val_loss: 0.5392 - val_acc: 0.7759\n",
      "Epoch 26/50\n",
      "513/513 [==============================] - 0s 406us/step - loss: 0.5174 - acc: 0.7349 - val_loss: 0.5359 - val_acc: 0.7586\n",
      "Epoch 27/50\n",
      "513/513 [==============================] - 0s 396us/step - loss: 0.5043 - acc: 0.7680 - val_loss: 0.5319 - val_acc: 0.7586\n",
      "Epoch 28/50\n",
      "513/513 [==============================] - 0s 411us/step - loss: 0.5046 - acc: 0.7661 - val_loss: 0.5113 - val_acc: 0.7759\n",
      "Epoch 29/50\n",
      "513/513 [==============================] - 0s 388us/step - loss: 0.5047 - acc: 0.7466 - val_loss: 0.5487 - val_acc: 0.7586\n",
      "Epoch 30/50\n",
      "513/513 [==============================] - 0s 383us/step - loss: 0.5049 - acc: 0.7524 - val_loss: 0.5152 - val_acc: 0.7759\n",
      "Epoch 31/50\n",
      "513/513 [==============================] - 0s 408us/step - loss: 0.4979 - acc: 0.7739 - val_loss: 0.5006 - val_acc: 0.7759\n",
      "Epoch 32/50\n",
      "513/513 [==============================] - 0s 404us/step - loss: 0.4975 - acc: 0.7739 - val_loss: 0.4977 - val_acc: 0.7759\n",
      "Epoch 33/50\n",
      "513/513 [==============================] - 0s 385us/step - loss: 0.4985 - acc: 0.7719 - val_loss: 0.5248 - val_acc: 0.7414\n",
      "Epoch 34/50\n",
      "513/513 [==============================] - 0s 390us/step - loss: 0.4938 - acc: 0.7758 - val_loss: 0.4962 - val_acc: 0.7759\n",
      "Epoch 35/50\n",
      "513/513 [==============================] - 0s 425us/step - loss: 0.4848 - acc: 0.7719 - val_loss: 0.4985 - val_acc: 0.7931\n",
      "Epoch 36/50\n",
      "513/513 [==============================] - 0s 523us/step - loss: 0.4872 - acc: 0.7914 - val_loss: 0.5065 - val_acc: 0.7759\n",
      "Epoch 37/50\n",
      "513/513 [==============================] - 0s 430us/step - loss: 0.4795 - acc: 0.7875 - val_loss: 0.4960 - val_acc: 0.7931\n",
      "Epoch 38/50\n",
      "513/513 [==============================] - 0s 373us/step - loss: 0.4813 - acc: 0.7797 - val_loss: 0.5091 - val_acc: 0.7414\n",
      "Epoch 39/50\n",
      "513/513 [==============================] - 0s 376us/step - loss: 0.4653 - acc: 0.7856 - val_loss: 0.5045 - val_acc: 0.7759\n",
      "Epoch 40/50\n",
      "513/513 [==============================] - 0s 385us/step - loss: 0.4686 - acc: 0.7836 - val_loss: 0.4733 - val_acc: 0.8103\n",
      "Epoch 41/50\n",
      "513/513 [==============================] - 0s 375us/step - loss: 0.4691 - acc: 0.8031 - val_loss: 0.4809 - val_acc: 0.8276\n",
      "Epoch 42/50\n",
      "513/513 [==============================] - 0s 399us/step - loss: 0.4708 - acc: 0.7934 - val_loss: 0.5288 - val_acc: 0.7586\n",
      "Epoch 43/50\n",
      "513/513 [==============================] - 0s 377us/step - loss: 0.4556 - acc: 0.7934 - val_loss: 0.5102 - val_acc: 0.7414\n",
      "Epoch 44/50\n",
      "513/513 [==============================] - 0s 414us/step - loss: 0.4472 - acc: 0.8109 - val_loss: 0.4770 - val_acc: 0.8103\n",
      "Epoch 45/50\n",
      "513/513 [==============================] - 0s 412us/step - loss: 0.4513 - acc: 0.8012 - val_loss: 0.4678 - val_acc: 0.8103\n",
      "Epoch 46/50\n",
      "513/513 [==============================] - 0s 411us/step - loss: 0.4712 - acc: 0.7953 - val_loss: 0.5044 - val_acc: 0.7759\n",
      "Epoch 47/50\n",
      "513/513 [==============================] - 0s 393us/step - loss: 0.4529 - acc: 0.7992 - val_loss: 0.4622 - val_acc: 0.7931\n",
      "Epoch 48/50\n",
      "513/513 [==============================] - 0s 407us/step - loss: 0.4479 - acc: 0.8031 - val_loss: 0.4874 - val_acc: 0.7931\n",
      "Epoch 49/50\n",
      "513/513 [==============================] - 0s 400us/step - loss: 0.4534 - acc: 0.8051 - val_loss: 0.4910 - val_acc: 0.7931\n",
      "Epoch 50/50\n",
      "513/513 [==============================] - 0s 379us/step - loss: 0.4633 - acc: 0.8012 - val_loss: 0.4673 - val_acc: 0.8276\n",
      "Time taken to train the model:  13.753248929977417 secs\n"
     ]
    }
   ],
   "source": [
    "t0 = time.time()\n",
    "history = model.fit(x_train, y_train, epochs=50, batch_size=8, validation_split=0.1)\n",
    "t1 = time.time()\n",
    "\n",
    "print(\"Time taken to train the model: \", t1-t0, \"secs\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "plt.title('Model accuracy')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Validation'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('Model Loss Plot')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Validation'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "143/143 [==============================] - 0s 70us/step\n"
     ]
    }
   ],
   "source": [
    "eval_val = model.evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model has 74.825175 accuracy on the test data\n"
     ]
    }
   ],
   "source": [
    "print('The model has %f accuracy on the test data' % (eval_val[1] * 100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = model.predict_classes(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "confusion_matrix_val = confusion_matrix(y_test, y_pred)\n",
    "labels = ['Not Survived', 'Survived']\n",
    "\n",
    "# Create the plot figure \n",
    "fig = plt.figure()\n",
    "\n",
    "# Comfusion Matrix subplot\n",
    "subplt = fig.add_subplot(111)\n",
    "csubplt = subplt.matshow(confusion_matrix_val)\n",
    "\n",
    "plt.title('Confusion matrix of the neural network classifier')\n",
    "fig.colorbar(csubplt)\n",
    "\n",
    "subplt.set_xticklabels([''] + labels)\n",
    "subplt.set_yticklabels([''] + labels)\n",
    "\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above plot, we see that the number of correct values predicted is more than the incorrect values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Genetic Algorithm Implementation to get an optimal neural network model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from random import randint, random\n",
    "import functools \n",
    "from operator import add\n",
    "import numpy as np\n",
    "\n",
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def createWeights(m,n):\n",
    "    return np.random.uniform(-1,1,(m,n))\n",
    "\n",
    "def createBiases(n):\n",
    "    return np.random.uniform(-1,1,n)\n",
    "\n",
    "def individual():\n",
    "    'Create a neural network model which is the member of the population.'\n",
    "    model = Sequential()\n",
    "    model.add(Dense(32, input_dim=7, activation='sigmoid', name='layer_1'))\n",
    "    model.add(Dropout(0.1))\n",
    "    model.add(Dense(1, activation='sigmoid', name='layer_2'))\n",
    "    #model.summary()\n",
    "    \n",
    "    model.compile(loss='binary_crossentropy', optimizer='rmsprop', metrics=['accuracy'])\n",
    "    \n",
    "    model.get_layer('layer_1').set_weights([createWeights(7,32), createBiases(32)])\n",
    "    model.get_layer('layer_2').set_weights([createWeights(32,1), createBiases(1)])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getChildWeights(father, mother, layer):\n",
    "    'Create a mix of weights from both parents'\n",
    "    fatherWeights = father.get_layer(layer).get_weights()[0]\n",
    "    motherWeights = mother.get_layer(layer).get_weights()[0]\n",
    "    \n",
    "    childWeights = np.zeros(fatherWeights.shape)\n",
    "    # Randomizing only the row values\n",
    "    indexStatus = np.random.randint(2, size=(fatherWeights.shape[0],))\n",
    "    \n",
    "    for index in range(fatherWeights.shape[0]):\n",
    "        if indexStatus[index] == 0:\n",
    "            childWeights[index] = fatherWeights[index]\n",
    "        else:\n",
    "            childWeights[index] = motherWeights[index]\n",
    "    \n",
    "    return childWeights\n",
    "    \n",
    "def getChildBias(father, mother, layer):\n",
    "    'Create a mix of biases from both parents'\n",
    "    father_bias = father.get_layer(layer).get_weights()[1]\n",
    "    mother_bias = mother.get_layer(layer).get_weights()[1]\n",
    "    \n",
    "    child_bias = np.zeros(father_bias.shape)\n",
    "    # Randomizing only the row values\n",
    "    indexStatus = np.random.randint(2, size=(father_bias.shape[0],))\n",
    "    \n",
    "    for index in range(father_bias.shape[0]):\n",
    "        if indexStatus[index] == 0:\n",
    "            child_bias[index] = father_bias[index]\n",
    "        else:\n",
    "            child_bias[index] = mother_bias[index]\n",
    "\n",
    "    return child_bias"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def child(father, mother):\n",
    "    'Create a new neural network model child which is the member of the population.'\n",
    "    model = Sequential()\n",
    "    model.add(Dense(32, input_dim=7, activation='sigmoid', name='layer_1'))\n",
    "    model.add(Dropout(0.1))\n",
    "    model.add(Dense(1, activation='sigmoid', name='layer_2'))\n",
    "    #model.summary()\n",
    "    \n",
    "    model.compile(loss='binary_crossentropy', optimizer='rmsprop', metrics=['accuracy'])\n",
    "    \n",
    "    # Cross Breed the weights and biases\n",
    "    model.get_layer('layer_1').set_weights([getChildWeights(father, mother, 'layer_1'), getChildBias(father, mother, 'layer_1')])\n",
    "    model.get_layer('layer_2').set_weights([getChildWeights(father, mother, 'layer_2'), getChildBias(father, mother, 'layer_2')])\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def mutateWeights(individual, layer):\n",
    "    'Randomly mutate the weights of an individual'\n",
    "    individualWeights = individual.get_layer(layer).get_weights()[0]\n",
    "    randomWeights = createWeights(individualWeights.shape[0], individualWeights.shape[1])\n",
    "    \n",
    "    mutateWeights = np.zeros(individualWeights.shape)\n",
    "    indexStatus = np.random.randint(2, size=(individualWeights.shape[0],))\n",
    "    \n",
    "    for index in range(individualWeights.shape[0]):\n",
    "        if indexStatus[index] == 0:\n",
    "            mutateWeights[index] = individualWeights[index]\n",
    "        else:\n",
    "            mutateWeights[index] = randomWeights[index]\n",
    "    \n",
    "    return mutateWeights\n",
    "    \n",
    "def mutatedBias(individual, layer):\n",
    "    'Randomly mutate the biases of an individual'\n",
    "    individual_bias = individual.get_layer(layer).get_weights()[1]\n",
    "    random_bias = createBiases(individual_bias.shape[0])\n",
    "    \n",
    "    mutated_bias = np.zeros(individual_bias.shape)\n",
    "    indexStatus = np.random.randint(2, size=(individual_bias.shape[0],))\n",
    "    \n",
    "    for index in range(individual_bias.shape[0]):\n",
    "        if indexStatus[index] == 0:\n",
    "            mutated_bias[index] = individual_bias[index]\n",
    "        else:\n",
    "            mutated_bias[index] = random_bias[index]\n",
    "\n",
    "    return mutated_bias"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def population(count):\n",
    "    'count: the number of individuals in the population'\n",
    "    return [ individual() for ind in range(count) ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fitness(individual, x, y):\n",
    "    'Determine the fitness of an individual. Higher is better.'\n",
    "    return individual.evaluate(x, y)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def grade(pop, x, y):\n",
    "    'Find average fitness for a population.'\n",
    "    summed = functools.reduce(add, (fitness(ind, x, y) for ind in pop))\n",
    "    return summed / (len(pop) * 1.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The evolve function uses the above functions to generates a new population with the retain, random selection and mutation criterias."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evolve(pop, dataset_x, dataset_y, retain=0.2, random_select=0.05, mutate=0.01):\n",
    "    graded = [ (fitness(ind, dataset_x, dataset_y), ind) for ind in pop]\n",
    "    graded = [ x[1] for x in sorted(graded, key=lambda tup: tup[0])]\n",
    "\n",
    "    # number of individuals to retain\n",
    "    retain_length = int(len(graded)*retain)\n",
    "    parents = graded[-retain_length:]\n",
    "    \n",
    "    # randomly pick individuals to retain\n",
    "    for individual in graded[:len(graded)-retain_length:]:\n",
    "        if random_select > random():\n",
    "            parents.append(individual)\n",
    "    \n",
    "    # mutate some individuals\n",
    "    for individual in parents:\n",
    "        if mutate > random():\n",
    "            individual.get_layer('layer_1').set_weights([mutateWeights(individual, 'layer_1'), mutatedBias(individual, 'layer_1')])\n",
    "            individual.get_layer('layer_2').set_weights([mutateWeights(individual, 'layer_2'), mutatedBias(individual, 'layer_2')])\n",
    "    \n",
    "    # Cross bread parents to create children with the count of the initial population size\n",
    "    parents_length = len(parents)\n",
    "    desired_length = len(pop) - parents_length\n",
    "    children = []\n",
    "    while len(children) < desired_length:\n",
    "        father = randint(0, parents_length-1)\n",
    "        mother = randint(0, parents_length-1)\n",
    "        if father != mother:\n",
    "            father = parents[father]\n",
    "            mother = parents[mother]\n",
    "            children.append(child(father, mother))        \n",
    "    parents.extend(children)\n",
    "\n",
    "    return parents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the initial population and calculate the average fitness of the population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n"
     ]
    }
   ],
   "source": [
    "p_count = 10\n",
    "pop = population(p_count)\n",
    "average_fitness_history = [grade(pop, x_train, y_train),]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Started running the 0 generation\n",
      "571/571 [==============================] - 0s 160us/step\n",
      "571/571 [==============================] - 0s 155us/step\n",
      "571/571 [==============================] - 0s 153us/step\n",
      "571/571 [==============================] - 0s 181us/step\n",
      "571/571 [==============================] - 0s 144us/step\n",
      "571/571 [==============================] - 0s 142us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 140us/step\n",
      "571/571 [==============================] - 0s 131us/step\n",
      "571/571 [==============================] - 0s 105us/step\n",
      "571/571 [==============================] - 0s 133us/step\n",
      "571/571 [==============================] - 0s 147us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 0 generation with mean accuracy of: 0.619089\n",
      "Started running the 1 generation\n",
      "571/571 [==============================] - 0s 119us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 117us/step\n",
      "571/571 [==============================] - 0s 163us/step\n",
      "571/571 [==============================] - 0s 146us/step\n",
      "571/571 [==============================] - 0s 130us/step\n",
      "571/571 [==============================] - 0s 139us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 166us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 1 generation with mean accuracy of: 0.650263\n",
      "Started running the 2 generation\n",
      "571/571 [==============================] - 0s 147us/step\n",
      "571/571 [==============================] - 0s 178us/step\n",
      "571/571 [==============================] - 0s 135us/step\n",
      "571/571 [==============================] - 0s 150us/step\n",
      "571/571 [==============================] - 0s 130us/step\n",
      "571/571 [==============================] - 0s 146us/step\n",
      "571/571 [==============================] - 0s 141us/step\n",
      "571/571 [==============================] - 0s 122us/step\n",
      "571/571 [==============================] - 0s 124us/step\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 0s 118us/step\n",
      "571/571 [==============================] - 0s 133us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 2 generation with mean accuracy of: 0.675306\n",
      "Started running the 3 generation\n",
      "571/571 [==============================] - 0s 130us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 138us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 135us/step\n",
      "571/571 [==============================] - 0s 132us/step\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 0s 134us/step\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 0s 132us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 132us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 148us/step\n",
      "571/571 [==============================] - 0s 165us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 3 generation with mean accuracy of: 0.681261\n",
      "Started running the 4 generation\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 131us/step\n",
      "571/571 [==============================] - 0s 131us/step\n",
      "571/571 [==============================] - 0s 133us/step\n",
      "571/571 [==============================] - 0s 131us/step\n",
      "571/571 [==============================] - 0s 131us/step\n",
      "571/571 [==============================] - 0s 135us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 156us/step\n",
      "571/571 [==============================] - 0s 130us/step\n",
      "571/571 [==============================] - 0s 147us/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "Completed running the 4 generation with mean accuracy of: 0.694571\n",
      "Started running the 5 generation\n",
      "571/571 [==============================] - 0s 118us/step\n",
      "571/571 [==============================] - 0s 194us/step\n",
      "571/571 [==============================] - 0s 179us/step\n",
      "571/571 [==============================] - 0s 139us/step\n",
      "571/571 [==============================] - 0s 133us/step\n",
      "571/571 [==============================] - 0s 137us/step\n",
      "571/571 [==============================] - 0s 124us/step\n",
      "571/571 [==============================] - 0s 115us/step\n",
      "571/571 [==============================] - 0s 119us/step\n",
      "571/571 [==============================] - 0s 119us/step\n",
      "571/571 [==============================] - 0s 123us/step\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 5 generation with mean accuracy of: 0.702452\n",
      "Started running the 6 generation\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 129us/step\n",
      "571/571 [==============================] - 0s 132us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 128us/step\n",
      "571/571 [==============================] - 0s 126us/step\n",
      "571/571 [==============================] - 0s 121us/step\n",
      "571/571 [==============================] - 0s 122us/step\n",
      "571/571 [==============================] - 0s 125us/step\n",
      "571/571 [==============================] - 0s 161us/step\n",
      "571/571 [==============================] - 0s 141us/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 6 generation with mean accuracy of: 0.707531\n",
      "Started running the 7 generation\n",
      "571/571 [==============================] - 0s 110us/step\n",
      "571/571 [==============================] - 0s 117us/step\n",
      "571/571 [==============================] - 0s 115us/step\n",
      "571/571 [==============================] - 0s 113us/step\n",
      "571/571 [==============================] - 0s 110us/step\n",
      "571/571 [==============================] - 0s 108us/step\n",
      "571/571 [==============================] - 0s 106us/step\n",
      "571/571 [==============================] - 0s 110us/step\n",
      "571/571 [==============================] - 0s 110us/step\n",
      "571/571 [==============================] - 0s 144us/step\n",
      "571/571 [==============================] - 0s 139us/step\n",
      "571/571 [==============================] - 0s 134us/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 3s 5ms/step\n",
      "571/571 [==============================] - 3s 5ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 7 generation with mean accuracy of: 0.707531\n",
      "Started running the 8 generation\n",
      "571/571 [==============================] - 0s 123us/step\n",
      "571/571 [==============================] - 0s 139us/step\n",
      "571/571 [==============================] - 0s 143us/step\n",
      "571/571 [==============================] - 0s 130us/step\n",
      "571/571 [==============================] - 0s 146us/step\n",
      "571/571 [==============================] - 0s 126us/step\n",
      "571/571 [==============================] - 0s 139us/step\n",
      "571/571 [==============================] - 0s 133us/step\n",
      "571/571 [==============================] - 0s 127us/step\n",
      "571/571 [==============================] - 0s 112us/step\n",
      "571/571 [==============================] - 0s 109us/step\n",
      "571/571 [==============================] - 0s 113us/step\n",
      "571/571 [==============================] - 0s 109us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "Completed running the 8 generation with mean accuracy of: 0.707531\n",
      "Started running the 9 generation\n",
      "571/571 [==============================] - 0s 117us/step\n",
      "571/571 [==============================] - 0s 112us/step\n",
      "571/571 [==============================] - 0s 111us/step\n",
      "571/571 [==============================] - 0s 109us/step\n",
      "571/571 [==============================] - 0s 110us/step\n",
      "571/571 [==============================] - 0s 109us/step\n",
      "571/571 [==============================] - 0s 108us/step\n",
      "571/571 [==============================] - 0s 106us/step\n",
      "571/571 [==============================] - 0s 105us/step\n",
      "571/571 [==============================] - 0s 104us/step\n",
      "571/571 [==============================] - 0s 107us/step\n",
      "571/571 [==============================] - 0s 116us/step\n",
      "571/571 [==============================] - 0s 120us/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 3ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "571/571 [==============================] - 2s 4ms/step\n",
      "Completed running the 9 generation with mean accuracy of: 0.707531\n"
     ]
    }
   ],
   "source": [
    "generationCount = 10\n",
    "for i in range(generationCount):\n",
    "    print('Started running the %d generation' % (i))\n",
    "    pop = evolve(pop, x_train, y_train)\n",
    "    fitness_val = grade(pop, x_train, y_train)\n",
    "    print('Completed running the %d generation with mean accuracy of: %f' % (i, fitness_val))\n",
    "    average_fitness_history.append(fitness_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(average_fitness_history)\n",
    "plt.title('GA Average Accuracy Plot')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Generation')\n",
    "plt.legend(['Train'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the average accuracy increases with increase in the generation count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "571/571 [==============================] - 0s 196us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.6451380888551421, 0.7075306471508982]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pop[0].evaluate(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "571/571 [==============================] - 0s 224us/step\n",
      "The model has 70.753065 accuracy on the train data using genetic algorithm optimized neural network\n"
     ]
    }
   ],
   "source": [
    "print('The model has %f accuracy on the train data using genetic algorithm optimized neural network' % (pop[0].evaluate(x_train, y_train)[1] * 100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "143/143 [==============================] - 0s 186us/step\n"
     ]
    }
   ],
   "source": [
    "eval_val_ga = pop[0].evaluate(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The model has 62.237762 accuracy on the test data using genetic algorithm optimized neural network\n"
     ]
    }
   ],
   "source": [
    "print('The model has %f accuracy on the test data using genetic algorithm optimized neural network' % (eval_val_ga[1] * 100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Predict the classes for test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_ga = pop[0].predict_classes(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "confusion_matrix_val_ga = confusion_matrix(y_test, y_pred_ga)\n",
    "labels = ['Not Survived', 'Survived']\n",
    "\n",
    "# Create the plot figure \n",
    "fig = plt.figure()\n",
    "\n",
    "# Confusion Matrix subplot\n",
    "subplt = fig.add_subplot(111)\n",
    "csubplt = subplt.matshow(confusion_matrix_val_ga)\n",
    "\n",
    "plt.title('Confusion matrix of the neural network classifier with GA')\n",
    "fig.colorbar(csubplt)\n",
    "\n",
    "subplt.set_xticklabels([''] + labels)\n",
    "subplt.set_yticklabels([''] + labels)\n",
    "\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusion\n",
    "\n",
    "We see that the overall accuracy of the models generated using genetic algorithm increases with more generations. We also note that the training accuracy of 70% is reached for the genetic based neural network compared to 80% reached using backpropogation algorithm.\n",
    "\n",
    "*Note:* We were limited by the resources to run for more number of generations in genetic algorithm and limited it to 10 generations."
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}