Predicting-house-prices-above-or-below-median.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Predicting house prices.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "17iNG3plAeZR5-4So6mejvsoxa93zd3D2",
      "authorship_tag": "ABX9TyMLk9+qjqJdy762BNwrBCQU",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/btkcodedev/ml-projects/blob/main/Predicting-house-prices-above-or-below-median.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AiVgBUHc3Lu1"
      },
      "source": [
        "Used Neural Network Within Keras For Building The ANN For Project"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C1LthljHjbQO"
      },
      "source": [
        "#Help Referenced: https://medium.com/intuitive-deep-learning/build-your-first-neural-network-to-predict-house-prices-with-keras-eb5db60232c\n",
        "import pandas as pd"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ntMZeqxwkaJX",
        "outputId": "c6ab9942-22f5-4528-b4ca-2e4783e48c96"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X2WiWJlrkbLM"
      },
      "source": [
        "df = pd.read_csv('/content/drive/MyDrive/Colab Notebooks/housepricedata.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 225
        },
        "id": "KR4PaD30k52O",
        "outputId": "a8cfb7ac-adee-43c2-995e-e5748ec78511"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>LotArea</th>\n",
              "      <th>OverallQual</th>\n",
              "      <th>OverallCond</th>\n",
              "      <th>TotalBsmtSF</th>\n",
              "      <th>FullBath</th>\n",
              "      <th>HalfBath</th>\n",
              "      <th>BedroomAbvGr</th>\n",
              "      <th>TotRmsAbvGrd</th>\n",
              "      <th>Fireplaces</th>\n",
              "      <th>GarageArea</th>\n",
              "      <th>AboveMedianPrice</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>8450</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>856</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>8</td>\n",
              "      <td>0</td>\n",
              "      <td>548</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>9600</td>\n",
              "      <td>6</td>\n",
              "      <td>8</td>\n",
              "      <td>1262</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>1</td>\n",
              "      <td>460</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>11250</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>920</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>1</td>\n",
              "      <td>608</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9550</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>756</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>642</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>14260</td>\n",
              "      <td>8</td>\n",
              "      <td>5</td>\n",
              "      <td>1145</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>1</td>\n",
              "      <td>836</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   LotArea  OverallQual  OverallCond  ...  Fireplaces  GarageArea  AboveMedianPrice\n",
              "0     8450            7            5  ...           0         548                 1\n",
              "1     9600            6            8  ...           1         460                 1\n",
              "2    11250            7            5  ...           1         608                 1\n",
              "3     9550            7            5  ...           1         642                 0\n",
              "4    14260            8            5  ...           1         836                 1\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 442
        },
        "id": "Ig8t8SAMlCvx",
        "outputId": "c3f60b68-8221-4a9b-83ff-1168b41cdf57"
      },
      "source": [
        "df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>LotArea</th>\n",
              "      <th>OverallQual</th>\n",
              "      <th>OverallCond</th>\n",
              "      <th>TotalBsmtSF</th>\n",
              "      <th>FullBath</th>\n",
              "      <th>HalfBath</th>\n",
              "      <th>BedroomAbvGr</th>\n",
              "      <th>TotRmsAbvGrd</th>\n",
              "      <th>Fireplaces</th>\n",
              "      <th>GarageArea</th>\n",
              "      <th>AboveMedianPrice</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>8450</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>856</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>8</td>\n",
              "      <td>0</td>\n",
              "      <td>548</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>9600</td>\n",
              "      <td>6</td>\n",
              "      <td>8</td>\n",
              "      <td>1262</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>1</td>\n",
              "      <td>460</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>11250</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>920</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>1</td>\n",
              "      <td>608</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9550</td>\n",
              "      <td>7</td>\n",
              "      <td>5</td>\n",
              "      <td>756</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>642</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>14260</td>\n",
              "      <td>8</td>\n",
              "      <td>5</td>\n",
              "      <td>1145</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>1</td>\n",
              "      <td>836</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1455</th>\n",
              "      <td>7917</td>\n",
              "      <td>6</td>\n",
              "      <td>5</td>\n",
              "      <td>953</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>460</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1456</th>\n",
              "      <td>13175</td>\n",
              "      <td>6</td>\n",
              "      <td>6</td>\n",
              "      <td>1542</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>2</td>\n",
              "      <td>500</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1457</th>\n",
              "      <td>9042</td>\n",
              "      <td>7</td>\n",
              "      <td>9</td>\n",
              "      <td>1152</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>252</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1458</th>\n",
              "      <td>9717</td>\n",
              "      <td>5</td>\n",
              "      <td>6</td>\n",
              "      <td>1078</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>240</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1459</th>\n",
              "      <td>9937</td>\n",
              "      <td>5</td>\n",
              "      <td>6</td>\n",
              "      <td>1256</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>0</td>\n",
              "      <td>276</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1460 rows × 11 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      LotArea  OverallQual  ...  GarageArea  AboveMedianPrice\n",
              "0        8450            7  ...         548                 1\n",
              "1        9600            6  ...         460                 1\n",
              "2       11250            7  ...         608                 1\n",
              "3        9550            7  ...         642                 0\n",
              "4       14260            8  ...         836                 1\n",
              "...       ...          ...  ...         ...               ...\n",
              "1455     7917            6  ...         460                 1\n",
              "1456    13175            6  ...         500                 1\n",
              "1457     9042            7  ...         252                 1\n",
              "1458     9717            5  ...         240                 0\n",
              "1459     9937            5  ...         276                 0\n",
              "\n",
              "[1460 rows x 11 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y6f_f-bmlGYu"
      },
      "source": [
        "dataset = df.values  #Make it an array"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kowFcyPJlMU0",
        "outputId": "42c3aaf2-e1fb-4a09-f680-4723272a6bf3"
      },
      "source": [
        "dataset"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 8450,     7,     5, ...,     0,   548,     1],\n",
              "       [ 9600,     6,     8, ...,     1,   460,     1],\n",
              "       [11250,     7,     5, ...,     1,   608,     1],\n",
              "       ...,\n",
              "       [ 9042,     7,     9, ...,     2,   252,     1],\n",
              "       [ 9717,     5,     6, ...,     0,   240,     0],\n",
              "       [ 9937,     5,     6, ...,     0,   276,     0]])"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3I0vK1pwlPnI",
        "outputId": "afab9f98-adde-4aae-f7a0-91f4ea667d94"
      },
      "source": [
        "len(df.columns) #For splitting purpose"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "11"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dRDA_b6clfu1"
      },
      "source": [
        "x = dataset[:,0:10] #Spilt to 'every row', 'columnrange(0 to 9)'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A3IJeKp2lwLo",
        "outputId": "9d1b413d-75dd-464e-d2d6-cf953e3c3cd0"
      },
      "source": [
        "x"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 8450,     7,     5, ...,     8,     0,   548],\n",
              "       [ 9600,     6,     8, ...,     6,     1,   460],\n",
              "       [11250,     7,     5, ...,     6,     1,   608],\n",
              "       ...,\n",
              "       [ 9042,     7,     9, ...,     9,     2,   252],\n",
              "       [ 9717,     5,     6, ...,     5,     0,   240],\n",
              "       [ 9937,     5,     6, ...,     6,     0,   276]])"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2UPCjjBilw2J"
      },
      "source": [
        "y = dataset[:,10] #Spilt to 'every row', 'column(10)'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DzddJv1-l5_2",
        "outputId": "180c27d9-1c1d-4d4d-daef-075a69c26387"
      },
      "source": [
        "y"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, ..., 1, 0, 0])"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LklEcEbOl9BC"
      },
      "source": [
        "# Preprocessing for cleansing\n",
        "from sklearn import preprocessing"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64p3x95Gm3IA"
      },
      "source": [
        "minmaxscaler = preprocessing.MinMaxScaler()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bsvuJHMQnKPR"
      },
      "source": [
        "xscale = minmaxscaler.fit_transform(x)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bihjZJsWnOrE",
        "outputId": "afd34e33-f0d1-4e02-d5dd-b9c3d783f138"
      },
      "source": [
        "xscale"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.0334198 , 0.66666667, 0.5       , ..., 0.5       , 0.        ,\n",
              "        0.3864598 ],\n",
              "       [0.03879502, 0.55555556, 0.875     , ..., 0.33333333, 0.33333333,\n",
              "        0.32440056],\n",
              "       [0.04650728, 0.66666667, 0.5       , ..., 0.33333333, 0.33333333,\n",
              "        0.42877292],\n",
              "       ...,\n",
              "       [0.03618687, 0.66666667, 1.        , ..., 0.58333333, 0.66666667,\n",
              "        0.17771509],\n",
              "       [0.03934189, 0.44444444, 0.625     , ..., 0.25      , 0.        ,\n",
              "        0.16925247],\n",
              "       [0.04037019, 0.44444444, 0.625     , ..., 0.33333333, 0.        ,\n",
              "        0.19464034]])"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kb7SFHFJnQGW"
      },
      "source": [
        "# Separating training and testing dataset\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TsZDLeSPnixz"
      },
      "source": [
        "#Separating training and testing dataset\n",
        "# Total dataset = 100% = 1500 items\n",
        "# Training = 70% = 1050 items\n",
        "# Testing & Validation = 30% = 450 items\n",
        "xtrain, xvaltest, ytrain, yvaltest = train_test_split(xscale, y, test_size = 0.3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TezfSoQmn1ky"
      },
      "source": [
        "# Separating validation and testing dataset\n",
        "# Validation = 15% = 225 items\n",
        "# Testing = 15% = 225 items\n",
        "xval, xtest, yval, ytest = train_test_split(xvaltest, yvaltest, test_size = 0.5)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qY1ejCbJoUMf",
        "outputId": "55132a02-5c79-4440-af5b-d239e1fce6de"
      },
      "source": [
        "print(xtrain.shape, xval.shape, xtest.shape, ytrain.shape, yval.shape, ytest.shape)\n",
        "# 1460 data points are available, other 40 are excluded during the preprocessing"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(1022, 10) (219, 10) (219, 10) (1022,) (219,) (219,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QQ8iFq6Bpagu"
      },
      "source": [
        "from keras.models import Sequential\n",
        "from keras.layers import Dense"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qcZ1j_slqaAe"
      },
      "source": [
        "model = Sequential([\n",
        "                    #Dense refers to fully connected layers\n",
        "                    Dense(32, activation=\"relu\", input_shape=(10,)),  #First Hidden Layer \n",
        "                    Dense(32, activation=\"relu\"), #Second Hidden Layer\n",
        "                    Dense(1, activation=\"sigmoid\"),\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yjWawY8nrH45"
      },
      "source": [
        "model.compile(\n",
        "    optimizer= \"sgd\", #‘sgd’ refers to stochastic gradient descent\n",
        "    loss = \"binary_crossentropy\", #The loss function for outputs that take the values 1 or 0 is called binary cross entropy\n",
        "    metrics = ['accuracy'] #Also track accuracy\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ynwG2VIjrxVT",
        "outputId": "d01aa6c3-32de-4d17-b4e5-f4787d88fa00"
      },
      "source": [
        "hist = model.fit(\n",
        "    xtrain, ytrain,\n",
        "    batch_size=32, epochs= 100,\n",
        "    validation_data=(xval, yval)\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "32/32 [==============================] - 1s 7ms/step - loss: 0.6952 - accuracy: 0.4961 - val_loss: 0.6897 - val_accuracy: 0.5662\n",
            "Epoch 2/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6878 - accuracy: 0.5558 - val_loss: 0.6823 - val_accuracy: 0.5799\n",
            "Epoch 3/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6820 - accuracy: 0.5245 - val_loss: 0.6761 - val_accuracy: 0.5799\n",
            "Epoch 4/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.6766 - accuracy: 0.5382 - val_loss: 0.6709 - val_accuracy: 0.5936\n",
            "Epoch 5/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6719 - accuracy: 0.5890 - val_loss: 0.6662 - val_accuracy: 0.6119\n",
            "Epoch 6/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6670 - accuracy: 0.6282 - val_loss: 0.6615 - val_accuracy: 0.6438\n",
            "Epoch 7/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.6621 - accuracy: 0.6350 - val_loss: 0.6569 - val_accuracy: 0.7032\n",
            "Epoch 8/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6570 - accuracy: 0.6888 - val_loss: 0.6520 - val_accuracy: 0.7215\n",
            "Epoch 9/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6519 - accuracy: 0.7280 - val_loss: 0.6470 - val_accuracy: 0.7260\n",
            "Epoch 10/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6467 - accuracy: 0.7094 - val_loss: 0.6423 - val_accuracy: 0.7763\n",
            "Epoch 11/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6412 - accuracy: 0.7534 - val_loss: 0.6370 - val_accuracy: 0.7945\n",
            "Epoch 12/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.6357 - accuracy: 0.7632 - val_loss: 0.6315 - val_accuracy: 0.7991\n",
            "Epoch 13/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.6296 - accuracy: 0.7593 - val_loss: 0.6261 - val_accuracy: 0.7991\n",
            "Epoch 14/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6233 - accuracy: 0.7935 - val_loss: 0.6197 - val_accuracy: 0.8037\n",
            "Epoch 15/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6165 - accuracy: 0.7886 - val_loss: 0.6136 - val_accuracy: 0.7991\n",
            "Epoch 16/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6093 - accuracy: 0.8102 - val_loss: 0.6069 - val_accuracy: 0.8037\n",
            "Epoch 17/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.6017 - accuracy: 0.8297 - val_loss: 0.5997 - val_accuracy: 0.8037\n",
            "Epoch 18/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5937 - accuracy: 0.8239 - val_loss: 0.5929 - val_accuracy: 0.8082\n",
            "Epoch 19/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.5854 - accuracy: 0.8454 - val_loss: 0.5844 - val_accuracy: 0.8128\n",
            "Epoch 20/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.5765 - accuracy: 0.8454 - val_loss: 0.5763 - val_accuracy: 0.8174\n",
            "Epoch 21/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5673 - accuracy: 0.8552 - val_loss: 0.5673 - val_accuracy: 0.8174\n",
            "Epoch 22/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5577 - accuracy: 0.8503 - val_loss: 0.5587 - val_accuracy: 0.8174\n",
            "Epoch 23/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5477 - accuracy: 0.8581 - val_loss: 0.5497 - val_accuracy: 0.8265\n",
            "Epoch 24/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5375 - accuracy: 0.8571 - val_loss: 0.5413 - val_accuracy: 0.8311\n",
            "Epoch 25/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5272 - accuracy: 0.8669 - val_loss: 0.5323 - val_accuracy: 0.8311\n",
            "Epoch 26/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.5168 - accuracy: 0.8708 - val_loss: 0.5225 - val_accuracy: 0.8311\n",
            "Epoch 27/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.5063 - accuracy: 0.8669 - val_loss: 0.5143 - val_accuracy: 0.8356\n",
            "Epoch 28/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.4959 - accuracy: 0.8679 - val_loss: 0.5050 - val_accuracy: 0.8311\n",
            "Epoch 29/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4855 - accuracy: 0.8718 - val_loss: 0.4958 - val_accuracy: 0.8356\n",
            "Epoch 30/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4748 - accuracy: 0.8796 - val_loss: 0.4854 - val_accuracy: 0.8311\n",
            "Epoch 31/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4648 - accuracy: 0.8718 - val_loss: 0.4780 - val_accuracy: 0.8311\n",
            "Epoch 32/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4548 - accuracy: 0.8787 - val_loss: 0.4684 - val_accuracy: 0.8219\n",
            "Epoch 33/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4454 - accuracy: 0.8767 - val_loss: 0.4614 - val_accuracy: 0.8265\n",
            "Epoch 34/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.4354 - accuracy: 0.8806 - val_loss: 0.4538 - val_accuracy: 0.8265\n",
            "Epoch 35/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4260 - accuracy: 0.8816 - val_loss: 0.4440 - val_accuracy: 0.8265\n",
            "Epoch 36/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.4176 - accuracy: 0.8816 - val_loss: 0.4391 - val_accuracy: 0.8265\n",
            "Epoch 37/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.4091 - accuracy: 0.8787 - val_loss: 0.4345 - val_accuracy: 0.8265\n",
            "Epoch 38/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.4009 - accuracy: 0.8796 - val_loss: 0.4246 - val_accuracy: 0.8265\n",
            "Epoch 39/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3937 - accuracy: 0.8816 - val_loss: 0.4195 - val_accuracy: 0.8311\n",
            "Epoch 40/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3863 - accuracy: 0.8816 - val_loss: 0.4145 - val_accuracy: 0.8311\n",
            "Epoch 41/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3795 - accuracy: 0.8806 - val_loss: 0.4095 - val_accuracy: 0.8311\n",
            "Epoch 42/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3731 - accuracy: 0.8816 - val_loss: 0.4034 - val_accuracy: 0.8311\n",
            "Epoch 43/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3668 - accuracy: 0.8777 - val_loss: 0.3973 - val_accuracy: 0.8311\n",
            "Epoch 44/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3611 - accuracy: 0.8826 - val_loss: 0.3924 - val_accuracy: 0.8311\n",
            "Epoch 45/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3557 - accuracy: 0.8806 - val_loss: 0.3899 - val_accuracy: 0.8265\n",
            "Epoch 46/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3506 - accuracy: 0.8836 - val_loss: 0.3879 - val_accuracy: 0.8265\n",
            "Epoch 47/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3458 - accuracy: 0.8855 - val_loss: 0.3829 - val_accuracy: 0.8265\n",
            "Epoch 48/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3412 - accuracy: 0.8806 - val_loss: 0.3809 - val_accuracy: 0.8311\n",
            "Epoch 49/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3370 - accuracy: 0.8836 - val_loss: 0.3773 - val_accuracy: 0.8311\n",
            "Epoch 50/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3325 - accuracy: 0.8836 - val_loss: 0.3715 - val_accuracy: 0.8219\n",
            "Epoch 51/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3290 - accuracy: 0.8806 - val_loss: 0.3694 - val_accuracy: 0.8265\n",
            "Epoch 52/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3255 - accuracy: 0.8806 - val_loss: 0.3718 - val_accuracy: 0.8493\n",
            "Epoch 53/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8836 - val_loss: 0.3670 - val_accuracy: 0.8447\n",
            "Epoch 54/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3193 - accuracy: 0.8875 - val_loss: 0.3637 - val_accuracy: 0.8402\n",
            "Epoch 55/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3161 - accuracy: 0.8914 - val_loss: 0.3597 - val_accuracy: 0.8265\n",
            "Epoch 56/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3129 - accuracy: 0.8836 - val_loss: 0.3560 - val_accuracy: 0.8265\n",
            "Epoch 57/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3107 - accuracy: 0.8865 - val_loss: 0.3623 - val_accuracy: 0.8493\n",
            "Epoch 58/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.3083 - accuracy: 0.8845 - val_loss: 0.3589 - val_accuracy: 0.8493\n",
            "Epoch 59/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3062 - accuracy: 0.8875 - val_loss: 0.3598 - val_accuracy: 0.8447\n",
            "Epoch 60/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3042 - accuracy: 0.8865 - val_loss: 0.3547 - val_accuracy: 0.8493\n",
            "Epoch 61/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3016 - accuracy: 0.8845 - val_loss: 0.3507 - val_accuracy: 0.8402\n",
            "Epoch 62/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.3000 - accuracy: 0.8855 - val_loss: 0.3518 - val_accuracy: 0.8493\n",
            "Epoch 63/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2984 - accuracy: 0.8894 - val_loss: 0.3526 - val_accuracy: 0.8447\n",
            "Epoch 64/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.2963 - accuracy: 0.8894 - val_loss: 0.3482 - val_accuracy: 0.8493\n",
            "Epoch 65/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2944 - accuracy: 0.8894 - val_loss: 0.3440 - val_accuracy: 0.8447\n",
            "Epoch 66/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2934 - accuracy: 0.8865 - val_loss: 0.3472 - val_accuracy: 0.8493\n",
            "Epoch 67/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2921 - accuracy: 0.8885 - val_loss: 0.3500 - val_accuracy: 0.8447\n",
            "Epoch 68/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2904 - accuracy: 0.8894 - val_loss: 0.3501 - val_accuracy: 0.8447\n",
            "Epoch 69/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2889 - accuracy: 0.8914 - val_loss: 0.3466 - val_accuracy: 0.8447\n",
            "Epoch 70/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2875 - accuracy: 0.8943 - val_loss: 0.3414 - val_accuracy: 0.8447\n",
            "Epoch 71/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2865 - accuracy: 0.8914 - val_loss: 0.3410 - val_accuracy: 0.8493\n",
            "Epoch 72/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2849 - accuracy: 0.8904 - val_loss: 0.3444 - val_accuracy: 0.8447\n",
            "Epoch 73/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2842 - accuracy: 0.8914 - val_loss: 0.3448 - val_accuracy: 0.8447\n",
            "Epoch 74/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2828 - accuracy: 0.8933 - val_loss: 0.3382 - val_accuracy: 0.8447\n",
            "Epoch 75/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2820 - accuracy: 0.8904 - val_loss: 0.3370 - val_accuracy: 0.8447\n",
            "Epoch 76/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2815 - accuracy: 0.8953 - val_loss: 0.3351 - val_accuracy: 0.8447\n",
            "Epoch 77/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2799 - accuracy: 0.8933 - val_loss: 0.3363 - val_accuracy: 0.8447\n",
            "Epoch 78/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.2790 - accuracy: 0.8933 - val_loss: 0.3344 - val_accuracy: 0.8447\n",
            "Epoch 79/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2779 - accuracy: 0.8933 - val_loss: 0.3368 - val_accuracy: 0.8447\n",
            "Epoch 80/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2768 - accuracy: 0.8943 - val_loss: 0.3397 - val_accuracy: 0.8447\n",
            "Epoch 81/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2765 - accuracy: 0.8953 - val_loss: 0.3343 - val_accuracy: 0.8447\n",
            "Epoch 82/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.2751 - accuracy: 0.8963 - val_loss: 0.3339 - val_accuracy: 0.8447\n",
            "Epoch 83/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2748 - accuracy: 0.8973 - val_loss: 0.3334 - val_accuracy: 0.8447\n",
            "Epoch 84/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2737 - accuracy: 0.8953 - val_loss: 0.3364 - val_accuracy: 0.8447\n",
            "Epoch 85/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2733 - accuracy: 0.8943 - val_loss: 0.3317 - val_accuracy: 0.8447\n",
            "Epoch 86/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2720 - accuracy: 0.8992 - val_loss: 0.3302 - val_accuracy: 0.8447\n",
            "Epoch 87/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2712 - accuracy: 0.8982 - val_loss: 0.3250 - val_accuracy: 0.8493\n",
            "Epoch 88/100\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.2706 - accuracy: 0.8963 - val_loss: 0.3257 - val_accuracy: 0.8493\n",
            "Epoch 89/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2702 - accuracy: 0.8973 - val_loss: 0.3274 - val_accuracy: 0.8447\n",
            "Epoch 90/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2693 - accuracy: 0.8963 - val_loss: 0.3228 - val_accuracy: 0.8539\n",
            "Epoch 91/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2702 - accuracy: 0.8953 - val_loss: 0.3270 - val_accuracy: 0.8447\n",
            "Epoch 92/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2680 - accuracy: 0.8992 - val_loss: 0.3239 - val_accuracy: 0.8493\n",
            "Epoch 93/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2674 - accuracy: 0.8973 - val_loss: 0.3215 - val_accuracy: 0.8493\n",
            "Epoch 94/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2670 - accuracy: 0.8953 - val_loss: 0.3296 - val_accuracy: 0.8584\n",
            "Epoch 95/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2665 - accuracy: 0.8982 - val_loss: 0.3244 - val_accuracy: 0.8539\n",
            "Epoch 96/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2659 - accuracy: 0.8973 - val_loss: 0.3250 - val_accuracy: 0.8539\n",
            "Epoch 97/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2649 - accuracy: 0.9012 - val_loss: 0.3196 - val_accuracy: 0.8539\n",
            "Epoch 98/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2644 - accuracy: 0.8982 - val_loss: 0.3204 - val_accuracy: 0.8539\n",
            "Epoch 99/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2635 - accuracy: 0.8982 - val_loss: 0.3160 - val_accuracy: 0.8539\n",
            "Epoch 100/100\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2627 - accuracy: 0.8973 - val_loss: 0.3151 - val_accuracy: 0.8584\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fmSSMIxSsMbA",
        "outputId": "f31db231-ef32-4d58-a906-155b95d1c2b7"
      },
      "source": [
        "loss, accuracy = model.evaluate(xtest, ytest)\n",
        "print(xtest.shape, ytest.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "7/7 [==============================] - 0s 2ms/step - loss: 0.3020 - accuracy: 0.8767\n",
            "(219, 10) (219,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JYLMGk-6tRuM",
        "outputId": "69edc086-b62c-4d15-913f-67eb0c41a1a6"
      },
      "source": [
        "print(loss, accuracy)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.3020112216472626 0.8767123222351074\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mYwfbfQ0tduU"
      },
      "source": [
        "# Finding wheather overfits or not\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "SKkOD_OluVBK",
        "outputId": "a030f3fe-b3b0-4683-bcb6-7e89d8103307"
      },
      "source": [
        "plt.title('Model Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "plt.plot(hist.history['loss'])\n",
        "plt.plot(hist.history['val_loss'])\n",
        "plt.legend(['Train', 'Val'], loc=\"upper right\")  #Training & validation loss\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3wU1frH8c+TXkhPKCmQ0HtRRAULxYKKIAICooLYlWvv92eveK3YuaBeCyAqKIoUUVAslFAEQu8ESAghvZfz+2MWDBAgQDaT7D7v12tfZmdmZ5+5y93vzjkz54gxBqWUUu7Lw+4ClFJK2UuDQCml3JwGgVJKuTkNAqWUcnMaBEop5eY0CJRSys1pECh1HCISLyJGRLyqsO0oEfm9JupSqjppECiXISLbRaRYRCKPWL7C8WUeb09lJxcoStU0DQLlarYBww8+EZEOQIB95ShV+2kQKFfzGXBDhecjgU8rbiAiISLyqYikicgOEfk/EfFwrPMUkVdFZL+IbAWuqOS1E0Vkr4jsFpHnRcTzdAoWkWgRmSEiB0Rks4jcUmFdNxFJFJFsEUkVkdcdy/1E5HMRSReRTBFZKiINTqcO5b40CJSrWQQEi0gbxxf0MODzI7Z5GwgBmgIXYgXHjY51twD9gC5AV2DwEa/9BCgFmju2uQS4+TRrngIkA9GO93tRRHo71r0FvGWMCQaaAVMdy0c6jiEOiABuBwpOsw7lpjQIlCs6eFZwMbAO2H1wRYVweMwYk2OM2Q68Blzv2OQa4E1jzC5jzAHgpQqvbQBcDtxrjMkzxuwD3nDs75SISBzQA3jEGFNojFkJTOCfs5oSoLmIRBpjco0xiyosjwCaG2PKjDHLjDHZp1qHcm8aBMoVfQZcC4ziiGYhIBLwBnZUWLYDiHH8HQ3sOmLdQU0cr93raI7JBD4E6p9GrdHAAWNMzjHquQloCax3NP/0cyz/DJgDTBGRPSLyioh4n0Ydyo1pECiXY4zZgdVpfDkw7YjV+7F+TTepsKwx/5w17MVqbqm47qBdQBEQaYwJdTyCjTHtTqPcPUC4iARVVo8xZpMxZjhW2IwFvhaRQGNMiTHmGWNMW6A7VnPWDSh1CjQIlKu6CehtjMmruNAYU4bVzv6CiASJSBPgfv7pR5gK3C0isSISBjxa4bV7gbnAayISLCIeItJMRC48ibp8HR29fiLih/WF/yfwkmNZR0ftnwOIyHUiEmWMKQcyHfsoF5FeItLB0dSVjRVu5SdRh1KHaBAol2SM2WKMSTzG6n8BecBW4HdgEvCRY91/sZpc/gaWc/QZxQ2AD7AWyAC+BhqdRGm5WJ26Bx+9sS53jcc6O5gOPGWMmefYvi+QJCK5WB3Hw4wxBUBDx3tnY/WD/IrVXKTUSROdmEYppdybnhEopZSb0yBQSik3p0GglFJuToNAKaXcXJ0bCTEyMtLEx8fbXYZSStUpy5Yt22+MiapsXZ0Lgvj4eBITj3VVoFJKqcqIyI5jrdOmIaWUcnMaBEop5eY0CJRSys05tY9ARPpi3RbvCUwwxrx8xPo3gF6OpwFAfWNMqDNrUkq5n5KSEpKTkyksLLS7FKfz8/MjNjYWb++qD0brtCBwDIb1LtaY8MnAUhGZYYxZe3AbY8x9Fbb/F9ZEH0opVa2Sk5MJCgoiPj4eEbG7HKcxxpCenk5ycjIJCQlVfp0zm4a6AZuNMVuNMcVYszANOM72w4HJTqxHKeWmCgsLiYiIcOkQABARIiIiTvrMx5lBEMPhE3wk889kG4dxDAWcAPxyjPW3OuZtTUxLS6v2QpVSrs/VQ+CgUznO2tJZPAz42jFW/FGMMeONMV2NMV2joiq9H+KEtu3PY+zs9ehoq0opdThnBsFuDp/pKZYKc8ceYRhObhb6aW0K7y/Ywsuz1zvzbZRS6ijp6el07tyZzp0707BhQ2JiYg49Ly4uPu5rExMTufvuu51anzOvGloKtBCRBKwAGIY1j+xhRKQ1EAb85cRauOX8phxI3c0Hv24lqp4vN5/f1Jlvp5RSh0RERLBy5UoAnn76aerVq8eDDz54aH1paSleXpV/HXft2pWuXbs6tT6nnREYY0qBMVizPa0DphpjkkTkWRHpX2HTYcAU4+Q2G/nzbR7ZNopRrUp4fuY6pq9IdubbKaXUcY0aNYrbb7+ds88+m4cffpglS5Zw7rnn0qVLF7p3786GDRsAWLBgAf369QOsEBk9ejQ9e/akadOmjBs3rlpqcep9BMaYH4Efj1j25BHPn3ZmDYe0vgL54y2eyvw/9sWP5cGvVlFcWs7Qsxqf+LVKKZfxzPdJrN2TXa37bBsdzFNXtjvp1yUnJ/Pnn3/i6elJdnY2CxcuxMvLi3nz5vH444/zzTffHPWa9evXM3/+fHJycmjVqhV33HHHSd0zUJk6N+jcKYtoBtd9jXzSj7d9nueupi/wyDer2Z6ez0OXtMLDwz2uKFBK1R5DhgzB09MTgKysLEaOHMmmTZsQEUpKSip9zRVXXIGvry++vr7Ur1+f1NRUYmNjT6sO9wkCgOguMPRzPL8YwnsxY3m267O8v2ALO9PzeXVIJ/x9PO2uUCnlZKfyy91ZAgMDD/39xBNP0KtXL6ZPn8727dvp2bNnpa/x9fU99LenpyelpaWnXUdtuXy05jTrBVePxyN5KU8deJRnLmrAj2v2Muj9P9l1IN/u6pRSbiorK4uYGOtWq08++aRG39v9ggCg/dUw7Atk31pGrrudLwbHsCsjn/7v/M4fm/fbXZ1Syg09/PDDPPbYY3Tp0qVafuWfDKlrN1h17drVVNvENDv+hEnDwCeQ3X0nMmpOCVvScrm7TwvG9GqOl6d75qRSrmbdunW0adPG7jJqTGXHKyLLjDGVXofq3t90TbrDjT+Chycx0wbwfY8tDOgUzZvzNjF0/CJtKlJKuQX3DgKAhu3h1l8h/jz8Zt3HG34TeGdwKzam5HDZWwv5dsWxboZWSinXoEEAEBgBI76GCx6ClZ/T76+hzBsWROuGQdz75Uru/3IluUU122anlFI1RYPgIA9P6P1/cMN3UFJAg6n9+LLVr9zbO4FvV+7minEL+XtXpt1VKqVUtdMgOFLTnnDHn9B+EJ6/vcy9O+9m+vAYSkrLGfT+n4z/bQvl5XWrg10ppY5Hg6Ay/qEw6L8waCKkbaTTD/346aK99GkdxYs/rmfUJ0vZn1tkd5VKKVUtNAiOp8NguOMPaNSRwJl38YHf2/zn8mgWb03n8rcWsnhrut0VKqXqgF69ejFnzpzDlr355pvccccdlW7fs2dPqu0y+SrQIDiR0DgY+T30eQrZMIshi4cw7/IcAn29uHbCYt5foE1FSqnjGz58OFOmTDls2ZQpUxg+fLhNFR1Og6AqPDzh/Pvh1gUQ1JC4ubcwp+lU+rcNYezs9dz0v6Vk5B1/cgmllPsaPHgwM2fOPDQJzfbt29mzZw+TJ0+ma9eutGvXjqeeesq2+txr0LnT1aAd3PwL/DoWn4Wv8XrEEvr0eo77f0un39u/8861XejSOMzuKpVSxzPrUUhZXb37bNgBLnv5mKvDw8Pp1q0bs2bNYsCAAUyZMoVrrrmGxx9/nPDwcMrKyujTpw+rVq2iY8eO1VtbFegZwcny8oE+T8DIGUhxLv2W3MD889cjGK758C8+/mObzouslDpKxeahg81CU6dO5YwzzqBLly4kJSWxdu1aW2rTM4JTlXAB3P4HfHcXMYueZn6LK7i/6Fae+X4tiTsyGDuoI/V89X9epWqd4/xyd6YBAwZw3333sXz5cvLz8wkPD+fVV19l6dKlhIWFMWrUKAoLC22pTc8ITkdgBAyfDJe8gPeWOYzL/hevdS9h1uq9DHjndzal5thdoVKqlqhXrx69evVi9OjRDB8+nOzsbAIDAwkJCSE1NZVZs2bZVpsGwekSge5j4MbZiDEMWjGaBWctJje/kKve/YO5SSl2V6iUqiWGDx/O33//zfDhw+nUqRNdunShdevWXHvttfTo0cO2utx7GOrqVpAJPz4Eq6dS3PAM/lV0B3P2BnL/xS0Z06u5ToeplE10GGodhrrmHLwjefBH+GRu4YP8+3my2RZe/2kjd01aTn6xDlynlKp9NAicof0guP0PJKIFo3c/wbct5zIvaTeD3/+L3ZkFdlenlFKH0SBwltA4GD0buo6m885PWNL4PbIPpDLgnd9J3H7A7uqUcjt1rRn8VJ3KcWoQOJOXL/R7Awa8R9j+ZfwS8iztvfdw7YTFzF6z1+7qlHIbfn5+pKenu3wYGGNIT0/Hz8/vpF6nncU1ZddS+HIEpiiXFwMeZMK+Vjw7oD3Xn9PE7sqUcnklJSUkJyfbdp1+TfLz8yM2NhZvb+/Dlh+vs1jveKopcWfBrQuQKdfy7z3P0LbBKO7/tpy07ELuu7glInpFkVLO4u3tTUJCgt1l1FraNFSTgqPhxlnQcRgDMz/h+6gPmfjLap75fq2OYKqUso0GQU3z9oeBH0Dfl2mX+yfzQ1/gxz9X8PA3qyjTMFBK2UCDwA4icM4dyPXTiCrfx9zQl/hr2QrunryCkrJyu6tTSrkZDQI7Ne2JXP8toeQwJ+Ql1q5ZrmGglKpxGgR2izsLRv5APc9SZga9yM6kRdwzRcNAKVVzNAhqg0Yd4cZZBPj5Mz3gBTKSfubeL1dSqmGglKoBGgS1RVRLuPknfCKa8LnvWGTNNB6btlqvJlJKOZ0GQW0SHA03/ohnXDfe9nkHs/ILnp+5zuXvhlRK2UuDoLbxD4Prp0GzXvzHezw5iz5m3M+b7a5KKeXCNAhqI29/ZNgkaNabsd7/JXn+eL5YvMPuqpRSLkqDoLY6IgyWzviQeWtT7a5KKeWCnBoEItJXRDaIyGYRefQY21wjImtFJElEJjmznjrH2w+PYZMwjXvwqvcHfDV5Ait2ZthdlVLKxTgtCETEE3gXuAxoCwwXkbZHbNMCeAzoYYxpB9zrrHrqLG8/PK+djGnQnnGeb/LOJ5+yIz3P7qqUUi7EmWcE3YDNxpitxphiYAow4IhtbgHeNcZkABhj9jmxnrrLLxjvG6YhoXG8Wf4SL02cTHZhid1VKaVchDODIAbYVeF5smNZRS2BliLyh4gsEpG+le1IRG4VkUQRSUxLS3NSubVcYCQ+o77DOzCcl/Oe5JWPp+oNZ0qpamF3Z7EX0ALoCQwH/isioUduZIwZb4zpaozpGhUVVcMl1iKhcfjd/CPe/kE8kPIw46d+Z3dFSikX4Mwg2A3EVXge61hWUTIwwxhTYozZBmzECgZ1LGHxBN46C0+fAIau/xcz5i2wuyKlVB3nzCBYCrQQkQQR8QGGATOO2OZbrLMBRCQSq6loqxNrcg3hTQm8bTbenp60XXgHiRv0HgOl1KlzWhAYY0qBMcAcYB0w1RiTJCLPikh/x2ZzgHQRWQvMBx4yxqQ7qyZX4hnZDI+hn5AgKeRMvpndGXolkVLq1Ojk9XVc2k9vEvXHU3zmP4LB972Nv4+n3SUppWqh401eb3dnsTpNURfdQ0r8VYzIn8Skzz60uxylVB2kQVDXidBwxAek1WvN0J3PMOeXn+2uSClVx2gQuAJvfyJv+YYSzwDa/XobG7dus7sipVQdokHgIjxDY2D4ZCIli8LPryU3TzuPlVJVo0HgQsJanMOu8/9Dx/K1rPjvnTqhjVKqSjQIXEyLPqNYHnMd52d+y6IfPrK7HKVUHaBB4II6jXqdzd6taJ/4b3ZuTrK7HKVULadB4II8vX0Juf4zEKFg8kiKigrsLkkpVYtpELioqMat2NZ9LK3KNrFy4t12l6OUqsU0CFxYx0tu4K+oIZy9byobftL+AqVU5TQIXFzn0e+w2rMNjf94lJwdq+wuRylVC2kQuDh/fz+8hn5KjvGn4PPhmIJMu0tSStUyGgRuoE3Llizs/CphxXvZ++lNoPcXKKUq0CBwE1cNGMwXQaOJ3juP7F/fsbscpVQtokHgJjw9hAtueIpfzJkELHgak7zM7pKUUrWEBoEbaVo/iD09XyfFhJL3xfWg/QVKKTQI3M61F3bi/Yh/45ufQsFXt0F5ud0lKaVspkHgZjw8hFtHDOU/5jr8t87GLHzV7pKUUjbTIHBDTSICibn0PqaVnQfzX4SNc+wuSSllIw0CN3X9ufFMi3mQ9aYJ5d/cDOlb7C5JKWUTDQI35eEhPD+4G2PK7ye/xGC+vA6K8+0uSyllAw0CNxYfGciISy/gzsI7kX1rYdbDdpeklLKBBoGbG9U9nvy4nvyXgbDiM/h7it0lKaVqmAaBm/PwEMYO7sjrpYPZ6NcRfrgP0jbYXZZSqgZpECiaRdVjTJ/WXJd5G0UefvDVjVBaZHdZSqkaokGgALj1gqZERTfh4ZLbYV+SdVmpUsotaBAoALw9PRg7qCM/FHZgUWg/+HMc7Fxkd1lKqRqgQaAOaR8Twm0XNOWmlIEUBkTD9NuhKNfuspRSTqZBoA5zd58WNIiM5MGS2zAZ2+GnJ+wuSSnlZBoE6jB+3p68PKgjP2Q3468GwyHxI1jxhd1lKaWcSINAHaVbQjjXn9OEG3ZeRk6jHvDDvbBrid1lKaWcRINAVerhvq2ICg5kVO6dmOAYmDICspLtLksp5QQaBKpSQX7evDCwPcvShM8TxkJJgRUGen+BUi5Hg0AdU+/WDejfKZpnF5exp/dbsHclzHva7rKUUtVMg0Ad15NXtiXQ14t/rWiE6XYrLHoPNv1kd1lKqWqkQaCOK7KeL0/2a8uyHRl8EXwz1G9n3V+Qk2p3aUqpauLUIBCRviKyQUQ2i8ijlawfJSJpIrLS8bjZmfWoUzOwSwwXtIzipbnbSL30PSjOhW9v1/mOlXIRTgsCEfEE3gUuA9oCw0WkbSWbfmmM6ex4THBWPerUiQgvXNUeAzzyWzHm0pdgyy/w1zt2l6aUqgbOPCPoBmw2xmw1xhQDU4ABTnw/5URx4QE8fGkrFmxIY7rHxdCmP/z8DCQvs7s0pdRpcmYQxAC7KjxPdiw70iARWSUiX4tIXGU7EpFbRSRRRBLT0tKcUauqguvPjefMJmE8O3Md+3u/CkGN4JvRUJhtd2lKqdNgd2fx90C8MaYj8BPwv8o2MsaMN8Z0NcZ0jYqKqtEC1T88PYSxgzqQX1TGU3N3w6CJkLnLmszGGLvLU0qdoioFgYgEioiH4++WItJfRLxP8LLdQMVf+LGOZYcYY9KNMQfvUJoAnFm1spVdmtcP4u4+zZm5ei9zc5pAr8dhzdeQNN3u0pRSp6iqZwS/AX4iEgPMBa4HPjnBa5YCLUQkQUR8gGHAjIobiEijCk/7A+uqWI+y0W0XNqN1wyCe+G4N2V3vgugu1sT3+QfsLk0pdQqqGgRijMkHrgbeM8YMAdod7wXGmFJgDDAH6wt+qjEmSUSeFZH+js3uFpEkEfkbuBsYdSoHoWqWt6cHLw/qSFpOEa/M3Qz934GCDJjzb7tLU0qdgioHgYicC4wAZjqWeZ7oRcaYH40xLY0xzYwxLziWPWmMmeH4+zFjTDtjTCdjTC9jzPpTOQhV8zrHhXJjjwQ+X7STpYXR0ONe+HsSbJ5nd2lKqZNU1SC4F3gMmO74Vd8UmO+8slRd8MAlLYkN8+eRb1ZR2P1+iGgB398HhVl2l6aUOglVCgJjzK/GmP7GmLGOTuP9xpi7nVybquUCfLx4YWAHtqbl8d7vu2HAu5CzxxqltKTQ7vKUUlVU1auGJolIsIgEAmuAtSLykHNLU3XBhS2juKpzNO8v2Mwm37Zw1fuwfSFMuxnKy+wuTylVBVVtGmprjMkGrgJmAQlYVw4pxf/1s0YofXTaasrbD4FLX4R138PMB/T+AqXqgKoGgbfjvoGrgBnGmBJA/x+uAGuE0n9f3oZlOzKYtGQnnHuX1Xm87GNYqsNHKVXbVTUIPgS2A4HAbyLSBNBxBdQhg8+MpXuzCMbOWk9KViFc9DQ06wNzn4D9m+wuTyl1HFXtLB5njIkxxlxuLDuAXk6uTdUhIsKLAztQUl7OY9NWWaeLA94Fbz/45mYoK7G7RKXUMVS1szhERF4/OPCbiLyGdXag1CHxkYE80rc18zek8dWyZAhuBFeOs6a4XPCy3eUppY6hqk1DHwE5wDWORzbwsbOKUnXXyHPj6ZYQznPfr2VPZgG07Q+dR8Dvr1tzGCilap2qBkEzY8xTjrkFthpjngGaOrMwVTd5eAivDu5EmTE88s0qjDHQ92WIam3dX7DjL7tLVEodoapBUCAi5x18IiI9gALnlKTqusYRATx2WWsWbtrPlKW7wC8YbvgOgqPhiyGwWyezUao2qWoQ3A68KyLbRWQ78A5wm9OqUnXeiLObcG7TCF6YuY7dmQVQrz7cMAMCwuGzqyFljd0lKqUcqnrV0N/GmE5AR6CjMaYL0Nuplak6zcNDeGVwR8qN4dGDTUQhMTDye/AOgC8GW5PaKKVsd1IzlBljsh13GAPc74R6lAuJCw/g8cvbsHDTfiYvcXzphzWB676G4nz4fJDOYaBULXA6U1VKtVWhXNaIsxvTo3kEL8xcy64D+dbCBu1g2BeQsQ2mXAsl2t2klJ1OJwh0iAl1QiLC2EEd8RDhvi9XUlpWbq1IOB8Gfgg7F1mzmymlbHPcIBCRHBHJruSRA0TXUI2qjosNC+D5ge1J3JHBO/M3/7Oi/dXQ4x5Y/ilsnGNfgUq5ueMGgTEmyBgTXMkjyBjjVVNFqrpvQOcYru4Sw7ifN5G4vUK/QK/HoX47+G4M5KXbV6BSbux0moaUOinPDGhHbFgA90xZSVaBY+whL1+4+kNrzuOZ9+mw1UrZQINA1ZggP2/eGtaZlOxCnpmR9M+Khh2sM4O138HiDzUMlKphGgSqRnVpHMZdPZsxbcVu5ial/LOixz3Q/GKY/QhMuwUKdZRzpWqKBoGqcWN6t6Bto2Aen76GjLxia6GHJ1z7JfT+P1jzDXx4gQ5FoVQN0SBQNc7Hy4PXrulEVkExT3xXYagJD0+44CEY9SOUFcOEi2H+SzqXgVJOpkGgbNGmUTD39GnBD6v28sOqPYevbHIu3PEHdBgMv74MEy6CtI32FKqUG9AgULa5/cJmdIoL5fFpq625CyryD4Orx8M1n0LWLvj8aijMsqdQpVycBoGyjZenB28N7UxpueG+L1dSVl7J1UJtB8DwLyF7N8x+vOaLVMoNaBAoW8VHBvJ0/3Ys3naA8b9trXyjuLPgvPtg5eewYVbNFqiUG9AgULYbcmYsV3RoxGtzN7AqObPyjS58BBq0hxl36x3ISlUzDQJlOxHhxYEdqB/kyx2fL+fAwUtKK/LyhYEfWHcgT7tZ+wuUqkYaBKpWCAnw5v3rziQtt4gxk5b/M0ppRQ07wBWvwtZf4cMLYe/fNV+oUi5Ig0DVGp3iQnnhqvb8uSWdl2atr3yjM0fBjT9CaZF1n0HiRzokhVKnSYNA1SpDusYxqns8E3/fxrTlyZVv1PgcuH0hxPeAH+6Db27SISmUOg0aBKrW+fcVbTg7IZzHpq0mac8x+gICI2HEN9DnSUj61hqSYs+Kmi1UKRehQaBqHW9PD9659gzCAny4/fNlZOZX0nkM4OEB5z9gNRWVFcPES3T0UqVOgQaBqpWignx577ozSMkq5N4vV1Je2c1mBzU+B27/HZr1saa9nHo9FBzjMlSl1FE0CFStdUbjMJ68sh0LNqTx5rwTjDUUEA7DJ8MlL1g3nX14AezfVDOFKlXHaRCoWu26sxsz5MxYxv2ymZmr9h5/YxHoPgZunA0l+VZTUXJizRSqVB3m1CAQkb4iskFENovIo8fZbpCIGBHp6sx6VN0jIjw/sD1nNgnjga9WsmZ3FW4kizsLRs8Bv2D435Ww6SfnF6pUHea0IBART+Bd4DKgLTBcRNpWsl0QcA+w2Fm1qLrN18uTD647k4hAX275NJF92YUnflFEMxg9FyKaw6Sh8PcU5xeqVB3lzDOCbsBmY8xWY0wxMAUYUMl2zwFjgSr8v1u5q6ggX8bfcCaZ+SXc/GkiuUWlJ35RUAMYNdO632D6bfDXu84vVKk6yJlBEAPsqvA82bHsEBE5A4gzxsw83o5E5FYRSRSRxLS0tOqvVNUJ7aJDeHt4F5L2ZHPbZ4kUlZad+EV+wTDia2jTH+Y8DvOehvJKhq9Qyo3Z1lksIh7A68ADJ9rWGDPeGNPVGNM1KirK+cWpWuuitg14ZVBH/ticzj2TV1Y+JtGRvHxhyCdw5o3w+xswaQjk6g8KpQ5yZhDsBuIqPI91LDsoCGgPLBCR7cA5wAztMFYnMujMWJ7s15bZSSk8Pn01pio3kHl4Qr834IrXYdtC+KAHbF3g9FqVqgucGQRLgRYikiAiPsAwYMbBlcaYLGNMpDEm3hgTDywC+htj9Ho/dUKjz0vg7t7NmZqYzCtzNlTtRSJw1k1wyy/gFwqfDoCP+sKyT/QGNOXWnBYExphSYAwwB1gHTDXGJInIsyLS31nvq9zHfRe35NqzG/P+gi1MWHiM2c0q07A93Dof+jwF+Qfg+3vg1Rbw+WBYOhGydp94H0q5EKnSaXUt0rVrV5OYqCcNylJWbhgzaTmz1qTwxtBODOwSe3I7MMYarG7117DhR8jYZi3vditc/Bx4+1V/0UrZQESWGWMqbXr3qulilKpOnh7CG0M7k5G/hAe/WoWflyeXdWhU9R2IQMwZ1uPSFyBtAyROhCXjYedfMPhjiGzhvANQqhbQISZUnefn7cmEkWfROS6Uf01ewew1Kae2IxGo3xou/w8M/9JqIvrwQpj1CGyZD6XHGAVVqTpOg0C5hHq+Xnxy41l0iA1hzKTl/LQ29fR22Kov3PEHNO8NiR/DZ1fBK03hx4f10lPlcrSPQLmU7MISrp+4hLV7shg3rMvJNRMdS3EebPsN1n4Hq6aCtz+cOwa6/wt8653+/pWqAcfrI9AzAuVSgv28+XR0NzrEhDBm8gqmrzjGdJcnwycQWl0GAz+AuxZD8z7w68swoQ9k7Dj9/StlMw0C5XJC/L357Kaz6RYfzv1T/2bS4p3Vt/PIFnDNp3DDd5Cz1woDHepa1XEaBMolBfp68fGNZ9GzZRSPT1/N+N+2VO8bNO0JN82zzq262sEAABWnSURBVBY+uQL+fPvw+w9Sk+Cnp2D+S1BSUL3vrVQ10z4C5dKKS8u5b+pKZq7ay129mvHgJa0Qkep7g7z98NUo2L7Qeh59hjV/cuoaEE8wZRDVBgZPhAbtqu99lTpJ2keg3JaPlwfjhnVheLc43p2/hSe/S6LsePMfn6zASBj1A4xJtO5UFg+rM/myV+DBjXDdN5CfDuN7wR/joERHW1e1j54RKLdgjOHl2ev58NetXNSmAW8N60ygbw3dT5mbBjPGwMbZUK8h9LgHzhwFPgE18/5KoWcESiEiPHZZG57p345f1qcy5IO/2JtVQ2339aJg+BQY+b3V2TznMRjX2Zo17Vg/xEqLYfPPUFpUMzUqt6ZBoNzKyO7xTBx1FjsP5DPgnT9YsTOjZt5YBBIusJqRbpwFIbHWrGkf9YW9qw7ftjgfpgyHz6+GcWdYN7SVlVjrinLgwNZjB4hSp0CbhpRb2pCSw82fLiU1q4jnrmrH0LMa12wB5eWw8gtrxrT8dOg8Ano9bt2gNmko7FwEFzxozZmQvBTqNYDyMsjfb72+xaXWfQ0B4TVbt6qzjtc0pEGg3FZGXjF3T1nBwk37ufbsxjx1ZVt8vTxrtoiCTFj4Kiz+0LrKKKghZCXDoP9Cu4HWL/9NP8GKz8A/DMITrGaj3/5jbTvkE4jVuZzUiWkQKHUMZeWG/8zZwAe/bqFTbAjvXHsGceE2dOJm7IBfnodNc2HQRGhx0fG3373Mumw1ey+c/wCcd691tZJSx6BBoNQJzF6TwkNf/Y2Hh/D6NZ3o06aBPYUYY/UnVEVBBsx8ENZ8DSGNrWG0I1vA/o3WI2MHZO2ybnSLOdNaHxjp3PpVraVBoFQV7EjP484vlpO0J5ubzkvg4b6tar6p6FRsWwizHoZ9aw9fHlgfQuOs/26eB37B1hDb7a4+dthk74VlH8OOP60+CVMG/uHQ425o0t35x6KcRoNAqSoqLCnjxR/X8elfO2jTKJhxwzrTokGQ3WWdWFkprP3WOqOIbAERzQ8fGTV1LXx3F+xZDq2ugMtfsa5cOmj3MvjrPWsf5WUQ3cUaPsPDE/atg9xUaH4R9H4CojvX/PGp06ZBoNRJ+mV9Kg99tYrcolIev7wN15/TBA+Pahyawg5lpbDoPZj/ovUF3+vf1rAXv78BW+eDbzB0uR663QzhTf95XXE+LP2vtV1BpvW68x8Aj2NcfZ65C3L3QVGWddlrwgXaf1ELaBAodQrScop46Ou/WbAhjfOaR/LK4I5Eh7rAF1rGDpj5AGz+yXoeWB+6j4Guo8H3OGc/hVlWn8TqqdDiEhj44eGXryYnwoKX/9nvQY27w4ipx9+3cjoNAqVOkTGGyUt28fzMtXh6CE/0a8uQM2Ord+A6OxhjDXmRfwDaDwJvv6q/LnEizHrUupw1qpV1JlGUbQ285x8O59wJDTtYfRL7N8IP91uXuI742lqmbKFBoNRp2pGex0NfrWLJ9gOcnRDOCwM70Ly+G89OlrwMfn/duhmuMBvKS6HTMOh2y9G//Nd+B1+Phkadod/rENUavHxP/j3LSmDdDOuei/wDMORjK3BUlWgQKFUNyssNUxN38dKs9eQXl3Jnz+bc1as5Pl46UssJrZ8JU0dCeQl4eEFkS2h8jtUBnXCBtSxtvdWp7RcMCRf+c/aQvsWaInT5p5CzB8ISoLTQ6q8Y8A50GHz8907fYl0F1bY/+IU4/1hrKQ0CparR/twinvthLd+t3EObRsG8NqQTbaO1yeOEMndB8hJr0p69q2DnX1Cca4WAKbceB3l4Qdw5UJJvXemEWJMBnX271T+RlwZfjbT20Wm4FSy+QeAXavVbBIRbfRqLx8OGHwFj3Wsx8H2IP8+e47eZBoFSTjA3KYXHp68hM7+Ym89vyh0XNiMkwNvusuqO0mLYtRi2/GJdxdSgvfXITbU6nDfPswKh/SDrERx99Ovn/p9130NZceXv4R8OZ90Esd1g9iNwYBucexecMdK6xPZYVz65IA0CpZwkI6+Y535Yy7QVuwny8+K2C5pyY4+EmpvrQFkd2KWFVl9FYRYUHLD6LsrLoMXF/1y6WpT7T3AA+IZA3FnQ8zHnjdeUk2Ld6Nest3P2fxI0CJRysrV7snn9pw3MW7ePiEAf7ujZjOvOaYKfdx24M9nd7N8MuxZZo7punGs1M136otXRfeTVYLlp1llLw/YQFn/4usIsK2yONQJsQSZMuAjSN1lnIJe9UvWrs5xAg0CpGrJiZwavzd3I75v30yDYl7t6NeearnEaCLVVQQZMv926lLb9IKuTOjcVsvfAriWwL8nazssfLn0eut5k9WUsnQg/PwtlRVYfRfe7IbL5P/stK4VJ18C2X6HjUGvI8egu1oCCHp6Qk2r9twZHjtUgUKqG/bUlndfmbiBxRwYRgT6M7B7P9ec0ISzQx+7S1JHKy61LYee/8E+HtX+YdblrwgXWgH1/vGn1ZTTrY90zkbzUau4JbQIrJ1l9FC0usS6hbXUZ/PwcLHoXrnzLmpZ0/Y/WRERF2Ye/95mj4LL/gJfz/11oEChlA2MMS7dn8OGvW/h5/T78vT0Z1i2Om85LIDZM5yuudbL3WPdD1Gtw9H0OxsDSCTD3CavPoe/L0PEaqykpdx8sGQ8rPoecveAdCCV51hVOl439Zx8HtsK6760O7KCG1g14f7xlXR019DMIjILCTGsWutDqnyhJg0Apm21IyWH8b1v5buVuDHBlx0bc0bM5rRrqsAt1Sk4KePmBf+jR68rLYPvv1j0PItDvTfA8wUUDa76Bb+8CT2/rbKQ411re5kq4/FUrMKqJBoFStcSezAIm/r6NyUt2kl9cxiVtG3Bnr+Z0jqvki0W5h72rrMEA/UKtEWGLsuH3N62O5UteqLYb4TQIlKplMvKK+eTP7Xzy53ayCkro3iyCO3s2p0fziLo/jpE6ffs3w4wx1g1zAAER1oiwPe6FNv1OaZcaBErVUrlFpUxevJMJv28lNbuIdtHBDOwSw+UdGrnGSKfq1JWXWzfVpa2z+hcObIVz7oJWfU9pdxoEStVyRaVlTF++m//9tYN1e60rS85oHEr/TtFc0TGaqKBTGKRNqQo0CJSqQ7btz+PH1Xv5/u89rE/JwUOgR/NI+rZvyMVtGlA/2L6bklTdpUGgVB21MTWHGSv38P2qPexIz0cEOseFclGbBvRuXZ/WDYO0T0FViW1BICJ9gbcAT2CCMeblI9bfDtwFlAG5wK3GmLVH7agCDQLljowxbEzNZW5SCj+tS2VVchYA0SF+XNk5mqFd42ga5cbzI6gTsiUIRMQT2AhcDCQDS4HhFb/oRSTYGJPt+Ls/cKcx5rg9IRoESsG+7ELmb9jH3KRUFmxMo6zc0C0hnEvbNeTshHDaNArGs67Psayq1fGCwJlDJHYDNhtjtjqKmAIMAA4FwcEQcAgE6lY7lVI2qR/sx9CzGjP0rMbsyy7k6+XJfL0smed+sP7vFeTnxRmNw6xHk1DObBJGgI+OiKoq58x/GTHArgrPk4Gzj9xIRO4C7gd8gErHahWRW4FbARo3rv5br5Wqy+oH+3Fnz+bc2bM5e7MKWLLtAIu3HWD5jgze/HkjxoCftwcXtozisvaN6NW6PiH+Om+C+oczm4YGA32NMTc7nl8PnG2MGXOM7a8FLjXGjDzefrVpSKmqyy4sYfmODH5Zv485SSmkZhfh4ehwPr9FFGfFh9M0KpCGwX54aFOSS7OraWg3EFfheaxj2bFMAd53Yj1KuZ1gP296tqpPz1b1efrKdqzYlcmvG/bx26b9vP3LJsodvwP9vD1oFlWPDjEhdIgNoVNsqPYzuBFnnhF4YXUW98EKgKXAtcaYpArbtDDGbHL8fSXw1LES6yA9I1CqemTmF7N2bzbb9uexLS2PDak5rN6dRWZ+CQDBfl50S4jgnKbhnNEkjHbRwfh66bwKdZUtZwTGmFIRGQPMwbp89CNjTJKIPAskGmNmAGNE5CKgBMgAjtsspJSqPqEBPnRvFkn3ZpGHlhljSM4oYNmODBZtTeevrenMW5cKgI+nB+1igukUG3rozKFJRICGgwvQG8qUUseVml3Iip0ZLN+ZyYqdGSTtySa/uAwAD4HoUH+aRtWja5Mwzm0WQafYUHy83GdS+LpC7yxWSlWbsnLD1rRckvZks3V/Htv357ExNYcNqTkYAz5eHkSH+BEV5Ev9YD9aNQiiQ2wIHWJCiKynYybZxa7OYqWUC/L0EFo0CKJFg8Mn1cnML2bxtgMs25HB3qxC0nIKSdqdxY+r93Lw92aQnxcxof7EhPpTP9iX8EAfIgJ9aV6/Hl0ahxLkp5e12kGDQClVLUIDfLi0XUMubXf4rFq5RaUk7c5i9e4sdh3IZ3dmIbszC1i1O4sDecWUOS5d8hBo0yiY9tEhNIkMID4ikPiIQJpGBeLnrf0QzqRBoJRyqnq+XpzdNIKzm0Ycta683JBZUELSniyWbs8gcfsBfl6fyv7c4kPbiEBsmD/NouodesSF+xPk5009Xy8i6/kQGuD8yd9dmQaBUso2Hh5CeKAP57eI4vwWUYeW5xaVsn1/Htv257ElLZfN+3LZmpbHoq3pFJaUH7WfmFB/2scE06phMMF+XtTz9SLY35tmUfVIiAzUzusT0CBQStU69Xy9aB8TQvuYw+fqLS837MkqYE9mIblFJeQUlpKSVciaPdmsTs5kTlLqUfvy8hAahwcQ7G+dQQT5edGifj3axYTQLjqY6BB/t7+rWoNAKVVneHgIsWEBxIYFVLq+tKycvOIy8opKOZBXzJa0XDam5rBtfx45haXkFZWyJ7OAOUkph+6q9vYUGoX4Ex3qR8NgP+oH+1E/yJcQf2+C/LwI8vOmQbAvsWEBh/oqsgtL2JmeT2iA9zFrqUs0CJRSLsPL04MQfw9C/L2JDvU/6ozioILiMtalZLN2TzbJGQXsybQey3ZmkJpdRHHp0c1PAA2CfSkuLSfDcfc1QPuYYPq2a8i5zSKoH2RdNlvXOrc1CJRSbsffx/PQMN1HMsaQXVBKdqHV9JRdWEJKViE7D+Sz80A+3p4eNIkIoHF4ALsO5DM7KYVX5248bB+BPp4E+VlnFMH+3oT6exMS4E14gI/j/gpfGgT7kRBpDfhn9yxzGgRKKVWBiBASYH1xV8VtFzYjJauQdXuzScspIi23iAN5xeQ4giSroISU7ELWp+RwIK+YgpKyw17v7+1J4/AAwgN9CA3wJjTAh9gwf2LD/IkLDyA21J/Ier5O7cfQIFBKqdPUMMSPhiF+J9zOGENuUSn7copIySq0Bvzbn8eO9HyyCorZvC+XA3nFpOcVH/Y6H08PGoX68cAlrejfKbra69cgUEqpGiIijiYj69LWHs0jK90uv7iU5IwCdh3IZ09mAcmZ1pVSEYHOuV9Cg0AppWqZAB8vWjYIouURw3g4i95loZRSbk6DQCml3JwGgVJKuTkNAqWUcnMaBEop5eY0CJRSys1pECillJvTIFBKKTdX5yavF5E0YMcpvjwS2F+N5dQV7njc7njM4J7H7Y7HDCd/3E2MMVGVrahzQXA6RCTRGNPV7jpqmjsetzseM7jncbvjMUP1Hrc2DSmllJvTIFBKKTfnbkEw3u4CbOKOx+2OxwzuedzueMxQjcftVn0ESimljuZuZwRKKaWOoEGglFJuzm2CQET6isgGEdksIo/aXY8ziEiciMwXkbUikiQi9ziWh4vITyKyyfHfo2fsruNExFNEVojID47nCSKy2PF5fykizpnayUYiEioiX4vIehFZJyLnuslnfZ/j3/caEZksIn6u9nmLyEcisk9E1lRYVulnK5ZxjmNfJSJnnOz7uUUQiIgn8C5wGdAWGC4ibe2tyilKgQeMMW2Bc4C7HMf5KPCzMaYF8LPjuau5B1hX4flY4A1jTHMgA7jJlqqc6y1gtjGmNdAJ6/hd+rMWkRjgbqCrMaY94AkMw/U+70+AvkcsO9ZnexnQwvG4FXj/ZN/MLYIA6AZsNsZsNcYUA1OAATbXVO2MMXuNMcsdf+dgfTHEYB3r/xyb/Q+4yp4KnUNEYoErgAmO5wL0Br52bOKKxxwCXABMBDDGFBtjMnHxz9rBC/AXES8gANiLi33expjfgANHLD7WZzsA+NRYFgGhItLoZN7PXYIgBthV4XmyY5nLEpF4oAuwGGhgjNnrWJUCNLCpLGd5E3gYKHc8jwAyjTGljueu+HknAGnAx44msQkiEoiLf9bGmN3Aq8BOrADIApbh+p83HPuzPe3vN3cJArciIvWAb4B7jTHZFdcZ63phl7lmWET6AfuMMcvsrqWGeQFnAO8bY7oAeRzRDORqnzWAo118AFYQRgOBHN2E4vKq+7N1lyDYDcRVeB7rWOZyRMQbKwS+MMZMcyxOPXiq6PjvPrvqc4IeQH8R2Y7V5Ncbq+081NF0AK75eScDycaYxY7nX2MFgyt/1gAXAduMMWnGmBJgGta/AVf/vOHYn+1pf7+5SxAsBVo4rizwwepcmmFzTdXO0TY+EVhnjHm9wqoZwEjH3yOB72q6NmcxxjxmjIk1xsRjfa6/GGNGAPOBwY7NXOqYAYwxKcAuEWnlWNQHWIsLf9YOO4FzRCTA8e/94HG79OftcKzPdgZwg+PqoXOArApNSFVjjHGLB3A5sBHYAvzb7nqcdIznYZ0urgJWOh6XY7WZ/wxsAuYB4XbX6qTj7wn84Pi7KbAE2Ax8BfjaXZ8TjrczkOj4vL8FwtzhswaeAdYDa4DPAF9X+7yByVh9ICVYZ383HeuzBQTrqsgtwGqsK6pO6v10iAmllHJz7tI0pJRS6hg0CJRSys1pECillJvTIFBKKTenQaCUUm5Og0CpI4hImYisrPCotoHbRCS+4oiSStUGXifeRCm3U2CM6Wx3EUrVFD0jUKqKRGS7iLwiIqtFZImINHcsjxeRXxxjwf8sIo0dyxuIyHQR+dvx6O7YlaeI/Ncxpv5cEfG37aCUQoNAqcr4H9E0NLTCuixjTAfgHaxRTwHeBv5njOkIfAGMcywfB/xqjOmENQ5QkmN5C+BdY0w7IBMY5OTjUeq49M5ipY4gIrnGmHqVLN8O9DbGbHUM7pdijIkQkf1AI2NMiWP5XmNMpIikAbHGmKIK+4gHfjLW5CKIyCOAtzHmeecfmVKV0zMCpU6OOcbfJ6Oowt9laF+dspkGgVInZ2iF//7l+PtPrJFPAUYACx1//wzcAYfmVA6pqSKVOhn6S0Spo/mLyMoKz2cbYw5eQhomIquwftUPdyz7F9ZMYQ9hzRp2o2P5PcB4EbkJ65f/HVgjSipVq2gfgVJV5Ogj6GqM2W93LUpVJ20aUkopN6dnBEop5eb0jEAppdycBoFSSrk5DQKllHJzGgRKKeXmNAiUUsrN/T9zEgsz/+LQPAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x8oVtYuIuV5u"
      },
      "source": [
        "plt.title('Model Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.plot(hist.history['accuracy'])\n",
        "plt.plot(hist.history['val_accuracy'])\n",
        "plt.legend(['Train', 'Val'], loc=\"upper right\")  #Training & validation accuracy\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NpQb7BKbvxmb"
      },
      "source": [
        "#Building Overfitting model\n",
        "modeltwo = Sequential([\n",
        "        Dense(1000, activation='relu', input_shape=(10,)),\n",
        "        Dense(1000, activation=\"relu\"),\n",
        "        Dense(1000, activation=\"relu\"),\n",
        "        Dense(1000, activation=\"relu\"),\n",
        "        Dense(1, activation=\"sigmoid\")\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "leh7v89uxoRW"
      },
      "source": [
        "modeltwo.compile(\n",
        "    optimizer='adam',\n",
        "    loss='binary_crossentropy',\n",
        "    metrics = ['accuracy']\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J-3Kj00nx6sy",
        "outputId": "1fdd922c-89f9-479e-90af-f55a4db24cf3"
      },
      "source": [
        "hist2 = modeltwo.fit(\n",
        "    xtrain, ytrain,\n",
        "    batch_size=32, epochs=100,\n",
        "    validation_data=(xval,yval)\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "32/32 [==============================] - 1s 40ms/step - loss: 0.1096 - accuracy: 0.9501 - val_loss: 0.4754 - val_accuracy: 0.8767\n",
            "Epoch 2/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0949 - accuracy: 0.9589 - val_loss: 0.4775 - val_accuracy: 0.8630\n",
            "Epoch 3/100\n",
            "32/32 [==============================] - 1s 43ms/step - loss: 0.1186 - accuracy: 0.9550 - val_loss: 0.4192 - val_accuracy: 0.8630\n",
            "Epoch 4/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.1152 - accuracy: 0.9521 - val_loss: 0.3150 - val_accuracy: 0.8858\n",
            "Epoch 5/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1523 - accuracy: 0.9315 - val_loss: 0.3776 - val_accuracy: 0.8813\n",
            "Epoch 6/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1156 - accuracy: 0.9560 - val_loss: 0.4573 - val_accuracy: 0.8858\n",
            "Epoch 7/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1146 - accuracy: 0.9530 - val_loss: 0.5460 - val_accuracy: 0.8721\n",
            "Epoch 8/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1152 - accuracy: 0.9540 - val_loss: 0.5930 - val_accuracy: 0.8584\n",
            "Epoch 9/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.1107 - accuracy: 0.9560 - val_loss: 0.4232 - val_accuracy: 0.8721\n",
            "Epoch 10/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0988 - accuracy: 0.9609 - val_loss: 0.4576 - val_accuracy: 0.8676\n",
            "Epoch 11/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1158 - accuracy: 0.9511 - val_loss: 0.4702 - val_accuracy: 0.8630\n",
            "Epoch 12/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0976 - accuracy: 0.9560 - val_loss: 0.5557 - val_accuracy: 0.8767\n",
            "Epoch 13/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.1086 - accuracy: 0.9569 - val_loss: 0.6413 - val_accuracy: 0.8447\n",
            "Epoch 14/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.1353 - accuracy: 0.9481 - val_loss: 0.4944 - val_accuracy: 0.8630\n",
            "Epoch 15/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0914 - accuracy: 0.9599 - val_loss: 0.5555 - val_accuracy: 0.8721\n",
            "Epoch 16/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0897 - accuracy: 0.9628 - val_loss: 0.6095 - val_accuracy: 0.8721\n",
            "Epoch 17/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0783 - accuracy: 0.9736 - val_loss: 0.6813 - val_accuracy: 0.8813\n",
            "Epoch 18/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0862 - accuracy: 0.9589 - val_loss: 0.6175 - val_accuracy: 0.8539\n",
            "Epoch 19/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.2319 - accuracy: 0.9051 - val_loss: 0.3743 - val_accuracy: 0.8630\n",
            "Epoch 20/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1315 - accuracy: 0.9442 - val_loss: 0.4677 - val_accuracy: 0.8676\n",
            "Epoch 21/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1024 - accuracy: 0.9589 - val_loss: 0.5349 - val_accuracy: 0.8767\n",
            "Epoch 22/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0776 - accuracy: 0.9687 - val_loss: 0.6192 - val_accuracy: 0.8676\n",
            "Epoch 23/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0893 - accuracy: 0.9628 - val_loss: 0.5976 - val_accuracy: 0.8630\n",
            "Epoch 24/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0879 - accuracy: 0.9658 - val_loss: 0.6082 - val_accuracy: 0.8721\n",
            "Epoch 25/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0941 - accuracy: 0.9579 - val_loss: 0.6135 - val_accuracy: 0.8447\n",
            "Epoch 26/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0885 - accuracy: 0.9609 - val_loss: 0.6426 - val_accuracy: 0.8721\n",
            "Epoch 27/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.0930 - accuracy: 0.9579 - val_loss: 0.6631 - val_accuracy: 0.8813\n",
            "Epoch 28/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0818 - accuracy: 0.9638 - val_loss: 0.6360 - val_accuracy: 0.8630\n",
            "Epoch 29/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0739 - accuracy: 0.9677 - val_loss: 0.6832 - val_accuracy: 0.8721\n",
            "Epoch 30/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0895 - accuracy: 0.9569 - val_loss: 0.7118 - val_accuracy: 0.8630\n",
            "Epoch 31/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0688 - accuracy: 0.9658 - val_loss: 0.7444 - val_accuracy: 0.8584\n",
            "Epoch 32/100\n",
            "32/32 [==============================] - 1s 41ms/step - loss: 0.1033 - accuracy: 0.9530 - val_loss: 0.6459 - val_accuracy: 0.8676\n",
            "Epoch 33/100\n",
            "32/32 [==============================] - 1s 40ms/step - loss: 0.1433 - accuracy: 0.9335 - val_loss: 0.5031 - val_accuracy: 0.8813\n",
            "Epoch 34/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.1113 - accuracy: 0.9501 - val_loss: 0.5798 - val_accuracy: 0.8767\n",
            "Epoch 35/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0853 - accuracy: 0.9638 - val_loss: 0.6206 - val_accuracy: 0.8721\n",
            "Epoch 36/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0683 - accuracy: 0.9716 - val_loss: 0.7174 - val_accuracy: 0.8630\n",
            "Epoch 37/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0636 - accuracy: 0.9746 - val_loss: 0.7874 - val_accuracy: 0.8584\n",
            "Epoch 38/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0598 - accuracy: 0.9736 - val_loss: 0.7557 - val_accuracy: 0.8721\n",
            "Epoch 39/100\n",
            "32/32 [==============================] - 1s 43ms/step - loss: 0.0646 - accuracy: 0.9677 - val_loss: 0.8403 - val_accuracy: 0.8767\n",
            "Epoch 40/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0571 - accuracy: 0.9746 - val_loss: 0.9890 - val_accuracy: 0.8630\n",
            "Epoch 41/100\n",
            "32/32 [==============================] - 1s 39ms/step - loss: 0.1036 - accuracy: 0.9560 - val_loss: 0.7900 - val_accuracy: 0.8539\n",
            "Epoch 42/100\n",
            "32/32 [==============================] - 1s 39ms/step - loss: 0.1687 - accuracy: 0.9256 - val_loss: 0.4229 - val_accuracy: 0.8721\n",
            "Epoch 43/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1083 - accuracy: 0.9521 - val_loss: 0.5462 - val_accuracy: 0.8721\n",
            "Epoch 44/100\n",
            "32/32 [==============================] - 1s 42ms/step - loss: 0.0899 - accuracy: 0.9658 - val_loss: 0.5727 - val_accuracy: 0.8721\n",
            "Epoch 45/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.1109 - accuracy: 0.9560 - val_loss: 0.4757 - val_accuracy: 0.8813\n",
            "Epoch 46/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0922 - accuracy: 0.9618 - val_loss: 0.6065 - val_accuracy: 0.8584\n",
            "Epoch 47/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.1025 - accuracy: 0.9560 - val_loss: 0.7137 - val_accuracy: 0.8813\n",
            "Epoch 48/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0722 - accuracy: 0.9716 - val_loss: 0.6576 - val_accuracy: 0.8676\n",
            "Epoch 49/100\n",
            "32/32 [==============================] - 1s 39ms/step - loss: 0.0567 - accuracy: 0.9736 - val_loss: 0.7088 - val_accuracy: 0.8676\n",
            "Epoch 50/100\n",
            "32/32 [==============================] - 1s 40ms/step - loss: 0.0516 - accuracy: 0.9795 - val_loss: 0.8216 - val_accuracy: 0.8584\n",
            "Epoch 51/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.1051 - accuracy: 0.9569 - val_loss: 0.5983 - val_accuracy: 0.8676\n",
            "Epoch 52/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1083 - accuracy: 0.9560 - val_loss: 0.5281 - val_accuracy: 0.8676\n",
            "Epoch 53/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0745 - accuracy: 0.9755 - val_loss: 0.6006 - val_accuracy: 0.8539\n",
            "Epoch 54/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0600 - accuracy: 0.9765 - val_loss: 0.6914 - val_accuracy: 0.8721\n",
            "Epoch 55/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0815 - accuracy: 0.9648 - val_loss: 0.7019 - val_accuracy: 0.8630\n",
            "Epoch 56/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0724 - accuracy: 0.9736 - val_loss: 0.6438 - val_accuracy: 0.8721\n",
            "Epoch 57/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0476 - accuracy: 0.9795 - val_loss: 0.8007 - val_accuracy: 0.8676\n",
            "Epoch 58/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0536 - accuracy: 0.9775 - val_loss: 0.7110 - val_accuracy: 0.8676\n",
            "Epoch 59/100\n",
            "32/32 [==============================] - 1s 41ms/step - loss: 0.0520 - accuracy: 0.9765 - val_loss: 0.8855 - val_accuracy: 0.8584\n",
            "Epoch 60/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.0732 - accuracy: 0.9648 - val_loss: 0.7212 - val_accuracy: 0.8584\n",
            "Epoch 61/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0712 - accuracy: 0.9687 - val_loss: 0.6884 - val_accuracy: 0.8630\n",
            "Epoch 62/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0863 - accuracy: 0.9658 - val_loss: 0.6663 - val_accuracy: 0.8676\n",
            "Epoch 63/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0598 - accuracy: 0.9775 - val_loss: 0.6564 - val_accuracy: 0.8721\n",
            "Epoch 64/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.0873 - accuracy: 0.9677 - val_loss: 0.6336 - val_accuracy: 0.8721\n",
            "Epoch 65/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.0889 - accuracy: 0.9628 - val_loss: 0.6802 - val_accuracy: 0.8493\n",
            "Epoch 66/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.0762 - accuracy: 0.9706 - val_loss: 0.7705 - val_accuracy: 0.8721\n",
            "Epoch 67/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.0667 - accuracy: 0.9726 - val_loss: 0.6662 - val_accuracy: 0.8447\n",
            "Epoch 68/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1015 - accuracy: 0.9579 - val_loss: 0.5635 - val_accuracy: 0.8721\n",
            "Epoch 69/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0650 - accuracy: 0.9726 - val_loss: 0.7568 - val_accuracy: 0.8767\n",
            "Epoch 70/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0700 - accuracy: 0.9667 - val_loss: 0.8531 - val_accuracy: 0.8493\n",
            "Epoch 71/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1220 - accuracy: 0.9667 - val_loss: 0.9030 - val_accuracy: 0.8721\n",
            "Epoch 72/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.1742 - accuracy: 0.9315 - val_loss: 0.4176 - val_accuracy: 0.8721\n",
            "Epoch 73/100\n",
            "32/32 [==============================] - 1s 42ms/step - loss: 0.0864 - accuracy: 0.9628 - val_loss: 0.5796 - val_accuracy: 0.8721\n",
            "Epoch 74/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.0631 - accuracy: 0.9765 - val_loss: 0.6149 - val_accuracy: 0.8858\n",
            "Epoch 75/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0588 - accuracy: 0.9746 - val_loss: 0.7860 - val_accuracy: 0.8904\n",
            "Epoch 76/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0573 - accuracy: 0.9775 - val_loss: 0.7917 - val_accuracy: 0.8630\n",
            "Epoch 77/100\n",
            "32/32 [==============================] - 1s 40ms/step - loss: 0.0455 - accuracy: 0.9785 - val_loss: 0.8210 - val_accuracy: 0.8721\n",
            "Epoch 78/100\n",
            "32/32 [==============================] - 1s 38ms/step - loss: 0.0512 - accuracy: 0.9755 - val_loss: 0.8799 - val_accuracy: 0.8676\n",
            "Epoch 79/100\n",
            "32/32 [==============================] - 1s 43ms/step - loss: 0.0617 - accuracy: 0.9716 - val_loss: 0.8066 - val_accuracy: 0.8767\n",
            "Epoch 80/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0401 - accuracy: 0.9834 - val_loss: 1.0038 - val_accuracy: 0.8539\n",
            "Epoch 81/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0466 - accuracy: 0.9824 - val_loss: 0.8675 - val_accuracy: 0.8676\n",
            "Epoch 82/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.0959 - accuracy: 0.9677 - val_loss: 0.7122 - val_accuracy: 0.8676\n",
            "Epoch 83/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0893 - accuracy: 0.9618 - val_loss: 0.7601 - val_accuracy: 0.8630\n",
            "Epoch 84/100\n",
            "32/32 [==============================] - 1s 43ms/step - loss: 0.0942 - accuracy: 0.9628 - val_loss: 0.7633 - val_accuracy: 0.8630\n",
            "Epoch 85/100\n",
            "32/32 [==============================] - 1s 42ms/step - loss: 0.0577 - accuracy: 0.9755 - val_loss: 0.7899 - val_accuracy: 0.8539\n",
            "Epoch 86/100\n",
            "32/32 [==============================] - 1s 41ms/step - loss: 0.0487 - accuracy: 0.9814 - val_loss: 0.7586 - val_accuracy: 0.8813\n",
            "Epoch 87/100\n",
            "32/32 [==============================] - 1s 39ms/step - loss: 0.0455 - accuracy: 0.9795 - val_loss: 0.9684 - val_accuracy: 0.8493\n",
            "Epoch 88/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0620 - accuracy: 0.9706 - val_loss: 1.0978 - val_accuracy: 0.8356\n",
            "Epoch 89/100\n",
            "32/32 [==============================] - 1s 35ms/step - loss: 0.0620 - accuracy: 0.9765 - val_loss: 0.7984 - val_accuracy: 0.8630\n",
            "Epoch 90/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0492 - accuracy: 0.9795 - val_loss: 0.8320 - val_accuracy: 0.8630\n",
            "Epoch 91/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0676 - accuracy: 0.9677 - val_loss: 0.8678 - val_accuracy: 0.8539\n",
            "Epoch 92/100\n",
            "32/32 [==============================] - 1s 41ms/step - loss: 0.0494 - accuracy: 0.9785 - val_loss: 0.9287 - val_accuracy: 0.8584\n",
            "Epoch 93/100\n",
            "32/32 [==============================] - 1s 43ms/step - loss: 0.0392 - accuracy: 0.9834 - val_loss: 1.0366 - val_accuracy: 0.8721\n",
            "Epoch 94/100\n",
            "32/32 [==============================] - 1s 37ms/step - loss: 0.0449 - accuracy: 0.9804 - val_loss: 1.0518 - val_accuracy: 0.8721\n",
            "Epoch 95/100\n",
            "32/32 [==============================] - 1s 41ms/step - loss: 0.0545 - accuracy: 0.9795 - val_loss: 1.0402 - val_accuracy: 0.8630\n",
            "Epoch 96/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0394 - accuracy: 0.9853 - val_loss: 1.1553 - val_accuracy: 0.8630\n",
            "Epoch 97/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0459 - accuracy: 0.9765 - val_loss: 1.0645 - val_accuracy: 0.8630\n",
            "Epoch 98/100\n",
            "32/32 [==============================] - 1s 36ms/step - loss: 0.0555 - accuracy: 0.9746 - val_loss: 1.0051 - val_accuracy: 0.8584\n",
            "Epoch 99/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.0395 - accuracy: 0.9814 - val_loss: 1.0282 - val_accuracy: 0.8539\n",
            "Epoch 100/100\n",
            "32/32 [==============================] - 1s 40ms/step - loss: 0.0595 - accuracy: 0.9775 - val_loss: 0.9753 - val_accuracy: 0.8539\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "k_LAIKDry7df",
        "outputId": "709a54d7-9b03-4fd7-f87c-940b4b1bb95e"
      },
      "source": [
        "plt.title('Model two loss')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.plot(hist2.history['loss'])\n",
        "plt.plot(hist2.history['val_loss'])\n",
        "plt.legend(['Train','Val'], loc=\"upper right\")\n",
        "plt.show() #Overfit Confirmed with inflexion point of Epoch 20"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "RHJOSDNOzZVA",
        "outputId": "33432c0e-f595-48d2-a2b1-ce128f72288c"
      },
      "source": [
        "plt.plot(hist2.history['accuracy'])\n",
        "plt.plot(hist2.history['val_accuracy'])\n",
        "plt.title('Model accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Val'], loc='lower right')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_L41P97e0ZRW"
      },
      "source": [
        "from keras.layers import Dropout\n",
        "from keras import regularizers"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U34YSe8q056o"
      },
      "source": [
        "model_3 = Sequential([\n",
        "    Dense(1000, activation='relu', kernel_regularizer=regularizers.l2(0.01), input_shape=(10,)), #Added L2 regularization\n",
        "    #This tells Keras to include the squared values of those parameters in our overall loss function, and weight them by 0.01 in the loss function.\n",
        "    Dropout(0.3), #neurons in the previous layer has a probability of 0.3 in dropping out during training\n",
        "    Dense(1000, activation='relu', kernel_regularizer=regularizers.l2(0.01)),\n",
        "    Dropout(0.3),\n",
        "    Dense(1000, activation='relu', kernel_regularizer=regularizers.l2(0.01)),\n",
        "    Dropout(0.3),\n",
        "    Dense(1000, activation='relu', kernel_regularizer=regularizers.l2(0.01)),\n",
        "    Dropout(0.3),\n",
        "    Dense(1, activation='sigmoid', kernel_regularizer=regularizers.l2(0.01)),\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iHCDA_wp2qgy",
        "outputId": "bdd9fdd6-bf0c-442a-9674-68d56052f7be"
      },
      "source": [
        "# model that survives overfittin using L2 regiularization and dropout\n",
        "model_3.compile(optimizer='adam',\n",
        "              loss='binary_crossentropy',\n",
        "              metrics=['accuracy'])\n",
        "hist_3 = model_3.fit(xtrain, ytrain,\n",
        "          batch_size=32, epochs=100,\n",
        "          validation_data=(xval, yval))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "32/32 [==============================] - 3s 53ms/step - loss: 14.2239 - accuracy: 0.6184 - val_loss: 3.8970 - val_accuracy: 0.7808\n",
            "Epoch 2/100\n",
            "32/32 [==============================] - 2s 61ms/step - loss: 1.6551 - accuracy: 0.8415 - val_loss: 0.6824 - val_accuracy: 0.8356\n",
            "Epoch 3/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.5455 - accuracy: 0.8748 - val_loss: 0.6231 - val_accuracy: 0.8174\n",
            "Epoch 4/100\n",
            "32/32 [==============================] - 2s 49ms/step - loss: 0.5046 - accuracy: 0.8738 - val_loss: 0.5600 - val_accuracy: 0.8356\n",
            "Epoch 5/100\n",
            "32/32 [==============================] - 2s 61ms/step - loss: 0.4780 - accuracy: 0.8777 - val_loss: 0.5340 - val_accuracy: 0.8311\n",
            "Epoch 6/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4737 - accuracy: 0.8845 - val_loss: 0.4940 - val_accuracy: 0.8539\n",
            "Epoch 7/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4878 - accuracy: 0.8679 - val_loss: 0.5353 - val_accuracy: 0.8447\n",
            "Epoch 8/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4641 - accuracy: 0.8806 - val_loss: 0.4941 - val_accuracy: 0.8539\n",
            "Epoch 9/100\n",
            "32/32 [==============================] - 2s 60ms/step - loss: 0.4454 - accuracy: 0.8816 - val_loss: 0.5117 - val_accuracy: 0.8539\n",
            "Epoch 10/100\n",
            "32/32 [==============================] - 2s 60ms/step - loss: 0.4404 - accuracy: 0.8885 - val_loss: 0.4627 - val_accuracy: 0.8721\n",
            "Epoch 11/100\n",
            "32/32 [==============================] - 2s 61ms/step - loss: 0.4447 - accuracy: 0.8865 - val_loss: 0.5272 - val_accuracy: 0.8539\n",
            "Epoch 12/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4831 - accuracy: 0.8650 - val_loss: 0.4961 - val_accuracy: 0.8311\n",
            "Epoch 13/100\n",
            "32/32 [==============================] - 2s 50ms/step - loss: 0.4394 - accuracy: 0.8894 - val_loss: 0.4514 - val_accuracy: 0.8721\n",
            "Epoch 14/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4441 - accuracy: 0.8816 - val_loss: 0.4877 - val_accuracy: 0.8265\n",
            "Epoch 15/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4737 - accuracy: 0.8699 - val_loss: 0.4676 - val_accuracy: 0.8630\n",
            "Epoch 16/100\n",
            "32/32 [==============================] - 2s 52ms/step - loss: 0.4280 - accuracy: 0.8885 - val_loss: 0.4495 - val_accuracy: 0.8584\n",
            "Epoch 17/100\n",
            "32/32 [==============================] - 1s 47ms/step - loss: 0.4445 - accuracy: 0.8855 - val_loss: 0.4496 - val_accuracy: 0.8584\n",
            "Epoch 18/100\n",
            "32/32 [==============================] - 2s 60ms/step - loss: 0.4386 - accuracy: 0.8836 - val_loss: 0.4511 - val_accuracy: 0.8721\n",
            "Epoch 19/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4301 - accuracy: 0.8904 - val_loss: 0.4484 - val_accuracy: 0.8721\n",
            "Epoch 20/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4550 - accuracy: 0.8689 - val_loss: 0.4545 - val_accuracy: 0.8721\n",
            "Epoch 21/100\n",
            "32/32 [==============================] - 2s 50ms/step - loss: 0.4356 - accuracy: 0.8894 - val_loss: 0.4406 - val_accuracy: 0.8630\n",
            "Epoch 22/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4795 - accuracy: 0.8601 - val_loss: 0.4724 - val_accuracy: 0.8311\n",
            "Epoch 23/100\n",
            "32/32 [==============================] - 2s 58ms/step - loss: 0.4396 - accuracy: 0.8826 - val_loss: 0.4542 - val_accuracy: 0.8676\n",
            "Epoch 24/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4280 - accuracy: 0.8875 - val_loss: 0.4730 - val_accuracy: 0.8356\n",
            "Epoch 25/100\n",
            "32/32 [==============================] - 2s 60ms/step - loss: 0.4385 - accuracy: 0.8816 - val_loss: 0.4434 - val_accuracy: 0.8539\n",
            "Epoch 26/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4714 - accuracy: 0.8669 - val_loss: 0.4578 - val_accuracy: 0.8767\n",
            "Epoch 27/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4205 - accuracy: 0.9022 - val_loss: 0.4547 - val_accuracy: 0.8630\n",
            "Epoch 28/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4187 - accuracy: 0.8894 - val_loss: 0.4659 - val_accuracy: 0.8584\n",
            "Epoch 29/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4410 - accuracy: 0.8855 - val_loss: 0.4658 - val_accuracy: 0.8356\n",
            "Epoch 30/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4347 - accuracy: 0.8875 - val_loss: 0.4678 - val_accuracy: 0.8676\n",
            "Epoch 31/100\n",
            "32/32 [==============================] - 2s 57ms/step - loss: 0.4252 - accuracy: 0.8982 - val_loss: 0.4544 - val_accuracy: 0.8630\n",
            "Epoch 32/100\n",
            "32/32 [==============================] - 1s 47ms/step - loss: 0.4396 - accuracy: 0.8826 - val_loss: 0.4792 - val_accuracy: 0.8676\n",
            "Epoch 33/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4350 - accuracy: 0.8826 - val_loss: 0.4381 - val_accuracy: 0.8721\n",
            "Epoch 34/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4195 - accuracy: 0.8924 - val_loss: 0.4830 - val_accuracy: 0.8676\n",
            "Epoch 35/100\n",
            "32/32 [==============================] - 2s 49ms/step - loss: 0.4445 - accuracy: 0.8669 - val_loss: 0.4392 - val_accuracy: 0.8584\n",
            "Epoch 36/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4261 - accuracy: 0.8875 - val_loss: 0.4433 - val_accuracy: 0.8402\n",
            "Epoch 37/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4311 - accuracy: 0.8914 - val_loss: 0.4360 - val_accuracy: 0.8676\n",
            "Epoch 38/100\n",
            "32/32 [==============================] - 2s 60ms/step - loss: 0.4337 - accuracy: 0.8836 - val_loss: 0.5205 - val_accuracy: 0.8493\n",
            "Epoch 39/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4304 - accuracy: 0.8845 - val_loss: 0.4591 - val_accuracy: 0.8676\n",
            "Epoch 40/100\n",
            "32/32 [==============================] - 2s 51ms/step - loss: 0.4290 - accuracy: 0.8914 - val_loss: 0.4575 - val_accuracy: 0.8584\n",
            "Epoch 41/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4353 - accuracy: 0.8914 - val_loss: 0.4575 - val_accuracy: 0.8539\n",
            "Epoch 42/100\n",
            "32/32 [==============================] - 1s 44ms/step - loss: 0.4314 - accuracy: 0.8865 - val_loss: 0.4361 - val_accuracy: 0.8539\n",
            "Epoch 43/100\n",
            "32/32 [==============================] - 2s 56ms/step - loss: 0.4188 - accuracy: 0.8924 - val_loss: 0.5773 - val_accuracy: 0.8219\n",
            "Epoch 44/100\n",
            "32/32 [==============================] - 2s 59ms/step - loss: 0.4410 - accuracy: 0.8845 - val_loss: 0.4417 - val_accuracy: 0.8676\n",
            "Epoch 45/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4287 - accuracy: 0.8904 - val_loss: 0.4357 - val_accuracy: 0.8630\n",
            "Epoch 46/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4298 - accuracy: 0.8865 - val_loss: 0.4362 - val_accuracy: 0.8676\n",
            "Epoch 47/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4154 - accuracy: 0.8865 - val_loss: 0.4319 - val_accuracy: 0.8721\n",
            "Epoch 48/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4325 - accuracy: 0.8933 - val_loss: 0.4344 - val_accuracy: 0.8584\n",
            "Epoch 49/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4170 - accuracy: 0.8855 - val_loss: 0.4697 - val_accuracy: 0.8721\n",
            "Epoch 50/100\n",
            "32/32 [==============================] - 2s 56ms/step - loss: 0.4198 - accuracy: 0.8904 - val_loss: 0.4420 - val_accuracy: 0.8402\n",
            "Epoch 51/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4327 - accuracy: 0.8894 - val_loss: 0.4384 - val_accuracy: 0.8493\n",
            "Epoch 52/100\n",
            "32/32 [==============================] - 2s 57ms/step - loss: 0.4415 - accuracy: 0.8748 - val_loss: 0.4486 - val_accuracy: 0.8493\n",
            "Epoch 53/100\n",
            "32/32 [==============================] - 2s 64ms/step - loss: 0.4346 - accuracy: 0.8836 - val_loss: 0.4799 - val_accuracy: 0.8721\n",
            "Epoch 54/100\n",
            "32/32 [==============================] - 2s 61ms/step - loss: 0.4246 - accuracy: 0.8894 - val_loss: 0.4342 - val_accuracy: 0.8721\n",
            "Epoch 55/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4273 - accuracy: 0.8914 - val_loss: 0.4417 - val_accuracy: 0.8767\n",
            "Epoch 56/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4232 - accuracy: 0.8855 - val_loss: 0.4313 - val_accuracy: 0.8539\n",
            "Epoch 57/100\n",
            "32/32 [==============================] - 1s 45ms/step - loss: 0.4137 - accuracy: 0.8982 - val_loss: 0.4346 - val_accuracy: 0.8539\n",
            "Epoch 58/100\n",
            "32/32 [==============================] - 2s 54ms/step - loss: 0.4195 - accuracy: 0.8865 - val_loss: 0.4853 - val_accuracy: 0.8676\n",
            "Epoch 59/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4294 - accuracy: 0.8894 - val_loss: 0.4417 - val_accuracy: 0.8721\n",
            "Epoch 60/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4298 - accuracy: 0.8836 - val_loss: 0.4347 - val_accuracy: 0.8813\n",
            "Epoch 61/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4255 - accuracy: 0.8826 - val_loss: 0.4451 - val_accuracy: 0.8584\n",
            "Epoch 62/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4354 - accuracy: 0.8816 - val_loss: 0.4377 - val_accuracy: 0.8767\n",
            "Epoch 63/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4255 - accuracy: 0.8875 - val_loss: 0.4379 - val_accuracy: 0.8767\n",
            "Epoch 64/100\n",
            "32/32 [==============================] - 1s 47ms/step - loss: 0.4194 - accuracy: 0.8894 - val_loss: 0.4602 - val_accuracy: 0.8402\n",
            "Epoch 65/100\n",
            "32/32 [==============================] - 2s 61ms/step - loss: 0.4389 - accuracy: 0.8816 - val_loss: 0.4884 - val_accuracy: 0.8584\n",
            "Epoch 66/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4322 - accuracy: 0.8826 - val_loss: 0.4297 - val_accuracy: 0.8630\n",
            "Epoch 67/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4284 - accuracy: 0.8777 - val_loss: 0.4579 - val_accuracy: 0.8311\n",
            "Epoch 68/100\n",
            "32/32 [==============================] - 2s 48ms/step - loss: 0.4273 - accuracy: 0.8816 - val_loss: 0.4329 - val_accuracy: 0.8813\n",
            "Epoch 69/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4396 - accuracy: 0.8816 - val_loss: 0.4361 - val_accuracy: 0.8813\n",
            "Epoch 70/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4162 - accuracy: 0.8943 - val_loss: 0.4433 - val_accuracy: 0.8676\n",
            "Epoch 71/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4202 - accuracy: 0.8894 - val_loss: 0.4319 - val_accuracy: 0.8767\n",
            "Epoch 72/100\n",
            "32/32 [==============================] - 2s 56ms/step - loss: 0.4233 - accuracy: 0.8914 - val_loss: 0.4346 - val_accuracy: 0.8767\n",
            "Epoch 73/100\n",
            "32/32 [==============================] - 2s 63ms/step - loss: 0.4283 - accuracy: 0.8904 - val_loss: 0.4313 - val_accuracy: 0.8539\n",
            "Epoch 74/100\n",
            "32/32 [==============================] - 2s 54ms/step - loss: 0.4246 - accuracy: 0.8836 - val_loss: 0.4502 - val_accuracy: 0.8584\n",
            "Epoch 75/100\n",
            "32/32 [==============================] - 1s 46ms/step - loss: 0.4234 - accuracy: 0.8943 - val_loss: 0.4387 - val_accuracy: 0.8676\n",
            "Epoch 76/100\n",
            "32/32 [==============================] - 2s 54ms/step - loss: 0.4387 - accuracy: 0.8855 - val_loss: 0.4432 - val_accuracy: 0.8767\n",
            "Epoch 77/100\n",
            "32/32 [==============================] - 2s 56ms/step - loss: 0.4256 - accuracy: 0.8855 - val_loss: 0.4312 - val_accuracy: 0.8676\n",
            "Epoch 78/100\n",
            "32/32 [==============================] - 2s 54ms/step - loss: 0.4254 - accuracy: 0.8924 - val_loss: 0.4311 - val_accuracy: 0.8539\n",
            "Epoch 79/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4299 - accuracy: 0.8806 - val_loss: 0.4328 - val_accuracy: 0.8676\n",
            "Epoch 80/100\n",
            "32/32 [==============================] - 1s 47ms/step - loss: 0.4338 - accuracy: 0.8826 - val_loss: 0.4692 - val_accuracy: 0.8767\n",
            "Epoch 81/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4090 - accuracy: 0.8973 - val_loss: 0.4452 - val_accuracy: 0.8630\n",
            "Epoch 82/100\n",
            "32/32 [==============================] - 2s 56ms/step - loss: 0.4183 - accuracy: 0.8845 - val_loss: 0.4358 - val_accuracy: 0.8721\n",
            "Epoch 83/100\n",
            "32/32 [==============================] - 2s 52ms/step - loss: 0.4217 - accuracy: 0.8924 - val_loss: 0.4305 - val_accuracy: 0.8584\n",
            "Epoch 84/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4241 - accuracy: 0.8875 - val_loss: 0.4343 - val_accuracy: 0.8767\n",
            "Epoch 85/100\n",
            "32/32 [==============================] - 2s 63ms/step - loss: 0.4549 - accuracy: 0.8875 - val_loss: 0.5101 - val_accuracy: 0.8493\n",
            "Epoch 86/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4445 - accuracy: 0.8836 - val_loss: 0.4598 - val_accuracy: 0.8721\n",
            "Epoch 87/100\n",
            "32/32 [==============================] - 2s 50ms/step - loss: 0.4184 - accuracy: 0.8963 - val_loss: 0.4754 - val_accuracy: 0.8721\n",
            "Epoch 88/100\n",
            "32/32 [==============================] - 2s 53ms/step - loss: 0.4212 - accuracy: 0.8865 - val_loss: 0.4310 - val_accuracy: 0.8767\n",
            "Epoch 89/100\n",
            "32/32 [==============================] - 2s 51ms/step - loss: 0.4364 - accuracy: 0.8787 - val_loss: 0.4511 - val_accuracy: 0.8402\n",
            "Epoch 90/100\n",
            "32/32 [==============================] - 2s 51ms/step - loss: 0.4233 - accuracy: 0.8904 - val_loss: 0.4326 - val_accuracy: 0.8767\n",
            "Epoch 91/100\n",
            "32/32 [==============================] - 2s 52ms/step - loss: 0.4197 - accuracy: 0.8855 - val_loss: 0.4325 - val_accuracy: 0.8584\n",
            "Epoch 92/100\n",
            "32/32 [==============================] - 1s 47ms/step - loss: 0.4213 - accuracy: 0.8796 - val_loss: 0.4612 - val_accuracy: 0.8311\n",
            "Epoch 93/100\n",
            "32/32 [==============================] - 2s 47ms/step - loss: 0.4417 - accuracy: 0.8748 - val_loss: 0.4824 - val_accuracy: 0.8402\n",
            "Epoch 94/100\n",
            "32/32 [==============================] - 2s 55ms/step - loss: 0.4536 - accuracy: 0.8650 - val_loss: 0.4362 - val_accuracy: 0.8676\n",
            "Epoch 95/100\n",
            "32/32 [==============================] - 2s 48ms/step - loss: 0.4153 - accuracy: 0.8777 - val_loss: 0.5184 - val_accuracy: 0.8539\n",
            "Epoch 96/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4232 - accuracy: 0.8933 - val_loss: 0.4310 - val_accuracy: 0.8813\n",
            "Epoch 97/100\n",
            "32/32 [==============================] - 2s 62ms/step - loss: 0.4208 - accuracy: 0.8845 - val_loss: 0.4290 - val_accuracy: 0.8584\n",
            "Epoch 98/100\n",
            "32/32 [==============================] - 2s 51ms/step - loss: 0.4336 - accuracy: 0.8875 - val_loss: 0.4400 - val_accuracy: 0.8721\n",
            "Epoch 99/100\n",
            "32/32 [==============================] - 2s 63ms/step - loss: 0.4268 - accuracy: 0.8943 - val_loss: 0.4350 - val_accuracy: 0.8813\n",
            "Epoch 100/100\n",
            "32/32 [==============================] - 2s 64ms/step - loss: 0.4292 - accuracy: 0.8806 - val_loss: 0.4563 - val_accuracy: 0.8767\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "8Wx6-R_D2rjc",
        "outputId": "911b94cc-6a82-439a-c0b0-36a3dea84a43"
      },
      "source": [
        "plt.plot(hist_3.history['loss'])\n",
        "plt.plot(hist_3.history['val_loss'])\n",
        "plt.title('Model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Val'], loc='upper right')\n",
        "plt.ylim(top=1.2, bottom=0) #window is zoomed in between 0 and 1.2 for the loss\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "LG0PLz923E0b",
        "outputId": "ea6fbf61-fc23-45a2-d475-fcf13a982598"
      },
      "source": [
        "plt.plot(hist_3.history['accuracy'])\n",
        "plt.plot(hist_3.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train', 'Val'], loc='upper right')\n",
        "plt.ylim(top=1.2, bottom=0) #window is zoomed in between 0 and 1.2 for the loss\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0cOiC5MY3wAr"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}