Breast Cancer Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 17024,
     "status": "ok",
     "timestamp": 1591293456346,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "SRUHBOqIQJ4_",
    "outputId": "617b66bc-2ca5-4c45-af44-8afc6eaae4b4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2020-06-04 17:57:20--  https://data.mendeley.com/datasets/wmy84gzngw/1/files/b63daee9-78de-4122-8475-9b3aa22ffd64/us-dataset.zip?dl=1\n",
      "Resolving data.mendeley.com (data.mendeley.com)... 162.159.133.86, 162.159.130.86, 2606:4700:7::a29f:8256, ...\n",
      "Connecting to data.mendeley.com (data.mendeley.com)|162.159.133.86|:443... connected.\n",
      "HTTP request sent, awaiting response... 302 Found\n",
      "Location: https://md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com/593779a5-2510-4e2d-adeb-71bdcf69a5f4 [following]\n",
      "--2020-06-04 17:57:21--  https://md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com/593779a5-2510-4e2d-adeb-71bdcf69a5f4\n",
      "Resolving md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com (md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com)... 52.218.53.211\n",
      "Connecting to md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com (md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com)|52.218.53.211|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 3264338 (3.1M) [application/zip]\n",
      "Saving to: ‘us-dataset.zip?dl=1.2’\n",
      "\n",
      "us-dataset.zip?dl=1 100%[===================>]   3.11M  1.44MB/s    in 2.2s    \n",
      "\n",
      "2020-06-04 17:57:25 (1.44 MB/s) - ‘us-dataset.zip?dl=1.2’ saved [3264338/3264338]\n",
      "\n",
      "Archive:  us-dataset.zip?dl=1\n",
      "   creating: originals/\n",
      "   creating: originals/benign/\n",
      "  inflating: originals/benign/us1.bmp  \n",
      "  inflating: originals/benign/us10.bmp  \n",
      "  inflating: originals/benign/us100.bmp  \n",
      "  inflating: originals/benign/us11.bmp  \n",
      "  inflating: originals/benign/us12.bmp  \n",
      "  inflating: originals/benign/us13.bmp  \n",
      "  inflating: originals/benign/us14.bmp  \n",
      "  inflating: originals/benign/us15.bmp  \n",
      "  inflating: originals/benign/us16.bmp  \n",
      "  inflating: originals/benign/us17.bmp  \n",
      "  inflating: originals/benign/us18.bmp  \n",
      "  inflating: originals/benign/us19.bmp  \n",
      "  inflating: originals/benign/us2.bmp  \n",
      "  inflating: originals/benign/us20.bmp  \n",
      "  inflating: originals/benign/us21.bmp  \n",
      "  inflating: originals/benign/us22.bmp  \n",
      "  inflating: originals/benign/us23.bmp  \n",
      "  inflating: originals/benign/us24.bmp  \n",
      "  inflating: originals/benign/us25.bmp  \n",
      "  inflating: originals/benign/us26.bmp  \n",
      "  inflating: originals/benign/us27.bmp  \n",
      "  inflating: originals/benign/us28.bmp  \n",
      "  inflating: originals/benign/us29.bmp  \n",
      "  inflating: originals/benign/us3.bmp  \n",
      "  inflating: originals/benign/us30.bmp  \n",
      "  inflating: originals/benign/us31.bmp  \n",
      "  inflating: originals/benign/us32.bmp  \n",
      "  inflating: originals/benign/us33.bmp  \n",
      "  inflating: originals/benign/us34.bmp  \n",
      "  inflating: originals/benign/us35.bmp  \n",
      "  inflating: originals/benign/us36.bmp  \n",
      "  inflating: originals/benign/us37.bmp  \n",
      "  inflating: originals/benign/us38.bmp  \n",
      "  inflating: originals/benign/us39.bmp  \n",
      "  inflating: originals/benign/us4.bmp  \n",
      "  inflating: originals/benign/us40.bmp  \n",
      "  inflating: originals/benign/us41.bmp  \n",
      "  inflating: originals/benign/us42.bmp  \n",
      "  inflating: originals/benign/us43.bmp  \n",
      "  inflating: originals/benign/us44.bmp  \n",
      "  inflating: originals/benign/us45.bmp  \n",
      "  inflating: originals/benign/us46.bmp  \n",
      "  inflating: originals/benign/us47.bmp  \n",
      "  inflating: originals/benign/us48.bmp  \n",
      "  inflating: originals/benign/us49.bmp  \n",
      "  inflating: originals/benign/us5.bmp  \n",
      "  inflating: originals/benign/us50.bmp  \n",
      "  inflating: originals/benign/us51.bmp  \n",
      "  inflating: originals/benign/us52.bmp  \n",
      "  inflating: originals/benign/us53.bmp  \n",
      "  inflating: originals/benign/us54.bmp  \n",
      "  inflating: originals/benign/us55.bmp  \n",
      "  inflating: originals/benign/us56.bmp  \n",
      "  inflating: originals/benign/us57.bmp  \n",
      "  inflating: originals/benign/us58.bmp  \n",
      "  inflating: originals/benign/us59.bmp  \n",
      "  inflating: originals/benign/us6.bmp  \n",
      "  inflating: originals/benign/us60.bmp  \n",
      "  inflating: originals/benign/us61.bmp  \n",
      "  inflating: originals/benign/us62.bmp  \n",
      "  inflating: originals/benign/us63.bmp  \n",
      "  inflating: originals/benign/us64.bmp  \n",
      "  inflating: originals/benign/us65.bmp  \n",
      "  inflating: originals/benign/us66.bmp  \n",
      "  inflating: originals/benign/us67.bmp  \n",
      "  inflating: originals/benign/us68.bmp  \n",
      "  inflating: originals/benign/us69.bmp  \n",
      "  inflating: originals/benign/us7.bmp  \n",
      "  inflating: originals/benign/us70.bmp  \n",
      "  inflating: originals/benign/us71.bmp  \n",
      "  inflating: originals/benign/us72.bmp  \n",
      "  inflating: originals/benign/us73.bmp  \n",
      "  inflating: originals/benign/us74.bmp  \n",
      "  inflating: originals/benign/us75.bmp  \n",
      "  inflating: originals/benign/us76.bmp  \n",
      "  inflating: originals/benign/us77.bmp  \n",
      "  inflating: originals/benign/us78.bmp  \n",
      "  inflating: originals/benign/us79.bmp  \n",
      "  inflating: originals/benign/us8.bmp  \n",
      "  inflating: originals/benign/us80.bmp  \n",
      "  inflating: originals/benign/us81.bmp  \n",
      "  inflating: originals/benign/us82.bmp  \n",
      "  inflating: originals/benign/us83.bmp  \n",
      "  inflating: originals/benign/us84.bmp  \n",
      "  inflating: originals/benign/us85.bmp  \n",
      "  inflating: originals/benign/us86.bmp  \n",
      "  inflating: originals/benign/us87.bmp  \n",
      "  inflating: originals/benign/us88.bmp  \n",
      "  inflating: originals/benign/us89.bmp  \n",
      "  inflating: originals/benign/us9.bmp  \n",
      "  inflating: originals/benign/us90.bmp  \n",
      "  inflating: originals/benign/us91.bmp  \n",
      "  inflating: originals/benign/us92.bmp  \n",
      "  inflating: originals/benign/us93.bmp  \n",
      "  inflating: originals/benign/us94.bmp  \n",
      "  inflating: originals/benign/us95.bmp  \n",
      "  inflating: originals/benign/us96.bmp  \n",
      "  inflating: originals/benign/us97.bmp  \n",
      "  inflating: originals/benign/us98.bmp  \n",
      "  inflating: originals/benign/us99.bmp  \n",
      "   creating: originals/malignant/\n",
      "  inflating: originals/malignant/us101.bmp  \n",
      "  inflating: originals/malignant/us102.bmp  \n",
      "  inflating: originals/malignant/us103.bmp  \n",
      "  inflating: originals/malignant/us104.bmp  \n",
      "  inflating: originals/malignant/us105.bmp  \n",
      "  inflating: originals/malignant/us106.bmp  \n",
      "  inflating: originals/malignant/us107.bmp  \n",
      "  inflating: originals/malignant/us108.bmp  \n",
      "  inflating: originals/malignant/us109.bmp  \n",
      "  inflating: originals/malignant/us110.bmp  \n",
      "  inflating: originals/malignant/us111.bmp  \n",
      "  inflating: originals/malignant/us112.bmp  \n",
      "  inflating: originals/malignant/us113.bmp  \n",
      "  inflating: originals/malignant/us114.bmp  \n",
      "  inflating: originals/malignant/us115.bmp  \n",
      "  inflating: originals/malignant/us116.bmp  \n",
      "  inflating: originals/malignant/us117.bmp  \n",
      "  inflating: originals/malignant/us118.bmp  \n",
      "  inflating: originals/malignant/us119.bmp  \n",
      "  inflating: originals/malignant/us120.bmp  \n",
      "  inflating: originals/malignant/us121.bmp  \n",
      "  inflating: originals/malignant/us122.bmp  \n",
      "  inflating: originals/malignant/us123.bmp  \n",
      "  inflating: originals/malignant/us124.bmp  \n",
      "  inflating: originals/malignant/us125.bmp  \n",
      "  inflating: originals/malignant/us126.bmp  \n",
      "  inflating: originals/malignant/us127.bmp  \n",
      "  inflating: originals/malignant/us128.bmp  \n",
      "  inflating: originals/malignant/us129.bmp  \n",
      "  inflating: originals/malignant/us130.bmp  \n",
      "  inflating: originals/malignant/us131.bmp  \n",
      "  inflating: originals/malignant/us132.bmp  \n",
      "  inflating: originals/malignant/us133.bmp  \n",
      "  inflating: originals/malignant/us134.bmp  \n",
      "  inflating: originals/malignant/us135.bmp  \n",
      "  inflating: originals/malignant/us136.bmp  \n",
      "  inflating: originals/malignant/us137.bmp  \n",
      "  inflating: originals/malignant/us138.bmp  \n",
      "  inflating: originals/malignant/us139.bmp  \n",
      "  inflating: originals/malignant/us140.bmp  \n",
      "  inflating: originals/malignant/us141.bmp  \n",
      "  inflating: originals/malignant/us142.bmp  \n",
      "  inflating: originals/malignant/us143.bmp  \n",
      "  inflating: originals/malignant/us144.bmp  \n",
      "  inflating: originals/malignant/us145.bmp  \n",
      "  inflating: originals/malignant/us146.bmp  \n",
      "  inflating: originals/malignant/us147.bmp  \n",
      "  inflating: originals/malignant/us148.bmp  \n",
      "  inflating: originals/malignant/us149.bmp  \n",
      "  inflating: originals/malignant/us150.bmp  \n",
      "  inflating: originals/malignant/us151.bmp  \n",
      "  inflating: originals/malignant/us152.bmp  \n",
      "  inflating: originals/malignant/us153.bmp  \n",
      "  inflating: originals/malignant/us154.bmp  \n",
      "  inflating: originals/malignant/us155.bmp  \n",
      "  inflating: originals/malignant/us156.bmp  \n",
      "  inflating: originals/malignant/us157.bmp  \n",
      "  inflating: originals/malignant/us158.bmp  \n",
      "  inflating: originals/malignant/us159.bmp  \n",
      "  inflating: originals/malignant/us160.bmp  \n",
      "  inflating: originals/malignant/us161.bmp  \n",
      "  inflating: originals/malignant/us162.bmp  \n",
      "  inflating: originals/malignant/us163.bmp  \n",
      "  inflating: originals/malignant/us164.bmp  \n",
      "  inflating: originals/malignant/us165.bmp  \n",
      "  inflating: originals/malignant/us166.bmp  \n",
      "  inflating: originals/malignant/us167.bmp  \n",
      "  inflating: originals/malignant/us168.bmp  \n",
      "  inflating: originals/malignant/us169.bmp  \n",
      "  inflating: originals/malignant/us170.bmp  \n",
      "  inflating: originals/malignant/us171.bmp  \n",
      "  inflating: originals/malignant/us172.bmp  \n",
      "  inflating: originals/malignant/us173.bmp  \n",
      "  inflating: originals/malignant/us174.bmp  \n",
      "  inflating: originals/malignant/us175.bmp  \n",
      "  inflating: originals/malignant/us176.bmp  \n",
      "  inflating: originals/malignant/us177.bmp  \n",
      "  inflating: originals/malignant/us178.bmp  \n",
      "  inflating: originals/malignant/us179.bmp  \n",
      "  inflating: originals/malignant/us180.bmp  \n",
      "  inflating: originals/malignant/us181.bmp  \n",
      "  inflating: originals/malignant/us182.bmp  \n",
      "  inflating: originals/malignant/us183.bmp  \n",
      "  inflating: originals/malignant/us184.bmp  \n",
      "  inflating: originals/malignant/us185.bmp  \n",
      "  inflating: originals/malignant/us186.bmp  \n",
      "  inflating: originals/malignant/us187.bmp  \n",
      "  inflating: originals/malignant/us188.bmp  \n",
      "  inflating: originals/malignant/us189.bmp  \n",
      "  inflating: originals/malignant/us190.bmp  \n",
      "  inflating: originals/malignant/us191.bmp  \n",
      "  inflating: originals/malignant/us192.bmp  \n",
      "  inflating: originals/malignant/us193.bmp  \n",
      "  inflating: originals/malignant/us194.bmp  \n",
      "  inflating: originals/malignant/us195.bmp  \n",
      "  inflating: originals/malignant/us196.bmp  \n",
      "  inflating: originals/malignant/us197.bmp  \n",
      "  inflating: originals/malignant/us198.bmp  \n",
      "  inflating: originals/malignant/us199.bmp  \n",
      "  inflating: originals/malignant/us200.bmp  \n",
      "  inflating: originals/malignant/us201.bmp  \n",
      "  inflating: originals/malignant/us202.bmp  \n",
      "  inflating: originals/malignant/us203.bmp  \n",
      "  inflating: originals/malignant/us204.bmp  \n",
      "  inflating: originals/malignant/us205.bmp  \n",
      "  inflating: originals/malignant/us206.bmp  \n",
      "  inflating: originals/malignant/us207.bmp  \n",
      "  inflating: originals/malignant/us208.bmp  \n",
      "  inflating: originals/malignant/us209.bmp  \n",
      "  inflating: originals/malignant/us210.bmp  \n",
      "  inflating: originals/malignant/us211.bmp  \n",
      "  inflating: originals/malignant/us212.bmp  \n",
      "  inflating: originals/malignant/us213.bmp  \n",
      "  inflating: originals/malignant/us214.bmp  \n",
      "  inflating: originals/malignant/us215.bmp  \n",
      "  inflating: originals/malignant/us216.bmp  \n",
      "  inflating: originals/malignant/us217.bmp  \n",
      "  inflating: originals/malignant/us218.bmp  \n",
      "  inflating: originals/malignant/us219.bmp  \n",
      "  inflating: originals/malignant/us220.bmp  \n",
      "  inflating: originals/malignant/us221.bmp  \n",
      "  inflating: originals/malignant/us222.bmp  \n",
      "  inflating: originals/malignant/us223.bmp  \n",
      "  inflating: originals/malignant/us224.bmp  \n",
      "  inflating: originals/malignant/us225.bmp  \n",
      "  inflating: originals/malignant/us226.bmp  \n",
      "  inflating: originals/malignant/us227.bmp  \n",
      "  inflating: originals/malignant/us228.bmp  \n",
      "  inflating: originals/malignant/us229.bmp  \n",
      "  inflating: originals/malignant/us230.bmp  \n",
      "  inflating: originals/malignant/us231.bmp  \n",
      "  inflating: originals/malignant/us232.bmp  \n",
      "  inflating: originals/malignant/us233.bmp  \n",
      "  inflating: originals/malignant/us234.bmp  \n",
      "  inflating: originals/malignant/us235.bmp  \n",
      "  inflating: originals/malignant/us236.bmp  \n",
      "  inflating: originals/malignant/us237.bmp  \n",
      "  inflating: originals/malignant/us238.bmp  \n",
      "  inflating: originals/malignant/us239.bmp  \n",
      "  inflating: originals/malignant/us240.bmp  \n",
      "  inflating: originals/malignant/us241.bmp  \n",
      "  inflating: originals/malignant/us242.bmp  \n",
      "  inflating: originals/malignant/us243.bmp  \n",
      "  inflating: originals/malignant/us244.bmp  \n",
      "  inflating: originals/malignant/us245.bmp  \n",
      "  inflating: originals/malignant/us246.bmp  \n",
      "  inflating: originals/malignant/us247.bmp  \n",
      "  inflating: originals/malignant/us248.bmp  \n",
      "  inflating: originals/malignant/us249.bmp  \n",
      "  inflating: originals/malignant/us250.bmp  \n"
     ]
    }
   ],
   "source": [
    "# Obtaining Dataset\n",
    "! wget https://data.mendeley.com/datasets/wmy84gzngw/1/files/b63daee9-78de-4122-8475-9b3aa22ffd64/us-dataset.zip?dl=1\n",
    "! unzip us-dataset.zip?dl=1\n",
    "! mv originals dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 33
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4828,
     "status": "ok",
     "timestamp": 1591293507787,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "zrTnWPlyUHyf",
    "outputId": "ce2d95e7-7967-4ca8-f34d-06eb0fc2d755"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# Importing Libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.layers import Activation, Dropout, Flatten, Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "6vFaRioRUdca"
   },
   "outputs": [],
   "source": [
    "# Dataset Preparation\n",
    "import os\n",
    "import random\n",
    "import shutil\n",
    "\n",
    "os.chdir('dataset/')\n",
    "os.mkdir('train/')\n",
    "os.mkdir('train/benign')\n",
    "os.mkdir('train/malignant')\n",
    "os.mkdir('validation/')\n",
    "os.mkdir('validation/benign')\n",
    "os.mkdir('validation/malignant')\n",
    "\n",
    "benign_samples = os.listdir('benign')\n",
    "random.shuffle(benign_samples)\n",
    "train_benign = benign_samples[:int(0.8*len(benign_samples))]\n",
    "val_benign = benign_samples[int(0.8*len(benign_samples)):]\n",
    "\n",
    "for f in train_benign:\n",
    "  shutil.move('benign/'+f, 'train/benign')\n",
    "for f in val_benign:\n",
    "  shutil.move('benign/'+f, 'validation/benign')\n",
    "os.rmdir('benign')\n",
    "\n",
    "malignant_samples = os.listdir('malignant')\n",
    "random.shuffle(malignant_samples)\n",
    "train_malignant = malignant_samples[:int(0.8*len(malignant_samples))]\n",
    "val_malignant = malignant_samples[int(0.8*len(malignant_samples)):]\n",
    "\n",
    "for f in train_malignant:\n",
    "  shutil.move('malignant/'+f, 'train/malignant')\n",
    "for f in val_malignant:\n",
    "  shutil.move('malignant/'+f, 'validation/malignant')\n",
    "os.rmdir('malignant')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 50
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 2648,
     "status": "ok",
     "timestamp": 1591293512957,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "0WDjslDjJHc_",
    "outputId": "26e3864b-80a5-4e15-b043-da740bc46ace"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 200 images belonging to 2 classes.\n",
      "Found 50 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "# Data Pre-Processing\n",
    "IMG_HEIGHT = 30\n",
    "IMG_WIDTH = 40\n",
    "\n",
    "datagen = ImageDataGenerator(rotation_range=40,\n",
    "                             width_shift_range=0.2,\n",
    "                             height_shift_range=0.2,\n",
    "                             shear_range=0.2,\n",
    "                             zoom_range=0.2,\n",
    "                             horizontal_flip=True,\n",
    "                             vertical_flip=True,\n",
    "                             fill_mode='nearest')\n",
    "\n",
    "training_set = datagen.flow_from_directory(\"train\",\n",
    "                                           target_size=(IMG_HEIGHT, IMG_WIDTH),\n",
    "                                           batch_size=32,\n",
    "                                           color_mode=\"grayscale\",\n",
    "                                           class_mode='binary')\n",
    "\n",
    "validation_set = datagen.flow_from_directory(\"validation\",\n",
    "                                             target_size=(IMG_HEIGHT, IMG_WIDTH),\n",
    "                                             batch_size=32,\n",
    "                                             color_mode=\"grayscale\",\n",
    "                                             class_mode='binary')\n",
    "label_map = training_set.class_indices #{'benign': 0, 'malignant': 1}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 302
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6428,
     "status": "ok",
     "timestamp": 1591293516941,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "tk1zU_trMHkO",
    "outputId": "2ab50b90-53e5-4af3-fe94-6287a294c396"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plotting 2x5 Images from Training Set\n",
    "sample_images, sample_labels = next(training_set)\n",
    "fig, axes = plt.subplots(2, 5, figsize=(10,5))\n",
    "axes = axes.flatten()\n",
    "for img, label, ax in zip(sample_images[:10], sample_labels[:10], axes):\n",
    "    ax.imshow(np.squeeze(img), cmap='gray')\n",
    "    ax.axis('off')\n",
    "    ax.set_title(label)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 537
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4455,
     "status": "ok",
     "timestamp": 1591293516942,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "h1QtxltoSHEn",
    "outputId": "2a70e4ac-59f8-42a0-c16a-219514198a8c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 28, 38, 64)        640       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 14, 19, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 12, 17, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 6, 8, 64)          0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 4, 6, 64)          36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 2, 3, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 384)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                24640     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 64)                4160      \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 103,361\n",
      "Trainable params: 103,361\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Deep Learning Model\n",
    "model = Sequential()\n",
    "model.add(Conv2D(64, (3, 3), input_shape=(IMG_HEIGHT, IMG_WIDTH, 1), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Conv2D(64, (3,3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "model.add(Conv2D(64, (3,3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(64, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(64, activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 992
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 317709,
     "status": "ok",
     "timestamp": 1591293836633,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "liQ_VUzBThb5",
    "outputId": "3dfde041-14b7-4468-8aa1-d1ccee0a1db0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "312/312 [==============================] - 12s 39ms/step - loss: 0.7873 - accuracy: 0.5711 - val_loss: 0.6522 - val_accuracy: 0.6304\n",
      "Epoch 2/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.6623 - accuracy: 0.6066 - val_loss: 0.6053 - val_accuracy: 0.6432\n",
      "Epoch 3/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.6388 - accuracy: 0.6297 - val_loss: 0.5936 - val_accuracy: 0.6611\n",
      "Epoch 4/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.6297 - accuracy: 0.6360 - val_loss: 0.7284 - val_accuracy: 0.6615\n",
      "Epoch 5/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.6027 - accuracy: 0.6576 - val_loss: 0.6184 - val_accuracy: 0.6854\n",
      "Epoch 6/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.5909 - accuracy: 0.6750 - val_loss: 0.5041 - val_accuracy: 0.6979\n",
      "Epoch 7/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.5583 - accuracy: 0.6944 - val_loss: 0.5618 - val_accuracy: 0.7583\n",
      "Epoch 8/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.5395 - accuracy: 0.7191 - val_loss: 0.4880 - val_accuracy: 0.7682\n",
      "Epoch 9/30\n",
      "312/312 [==============================] - 11s 34ms/step - loss: 0.5154 - accuracy: 0.7383 - val_loss: 0.6038 - val_accuracy: 0.7404\n",
      "Epoch 10/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.4750 - accuracy: 0.7652 - val_loss: 0.3939 - val_accuracy: 0.7539\n",
      "Epoch 11/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.4138 - accuracy: 0.8077 - val_loss: 0.4233 - val_accuracy: 0.7954\n",
      "Epoch 12/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3794 - accuracy: 0.8259 - val_loss: 0.4190 - val_accuracy: 0.7747\n",
      "Epoch 13/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3713 - accuracy: 0.8356 - val_loss: 0.4115 - val_accuracy: 0.8171\n",
      "Epoch 14/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3385 - accuracy: 0.8506 - val_loss: 0.3185 - val_accuracy: 0.7904\n",
      "Epoch 15/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3382 - accuracy: 0.8460 - val_loss: 0.4495 - val_accuracy: 0.8235\n",
      "Epoch 16/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3144 - accuracy: 0.8640 - val_loss: 0.3512 - val_accuracy: 0.8034\n",
      "Epoch 17/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.3150 - accuracy: 0.8666 - val_loss: 0.3465 - val_accuracy: 0.8171\n",
      "Epoch 18/30\n",
      "312/312 [==============================] - 10s 34ms/step - loss: 0.2840 - accuracy: 0.8744 - val_loss: 0.3768 - val_accuracy: 0.8294\n",
      "Epoch 19/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2708 - accuracy: 0.8824 - val_loss: 0.2862 - val_accuracy: 0.8478\n",
      "Epoch 20/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2743 - accuracy: 0.8825 - val_loss: 0.2557 - val_accuracy: 0.8568\n",
      "Epoch 21/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2515 - accuracy: 0.8905 - val_loss: 0.4176 - val_accuracy: 0.8286\n",
      "Epoch 22/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2469 - accuracy: 0.8958 - val_loss: 0.1861 - val_accuracy: 0.8711\n",
      "Epoch 23/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2408 - accuracy: 0.8975 - val_loss: 0.2714 - val_accuracy: 0.8529\n",
      "Epoch 24/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2367 - accuracy: 0.8975 - val_loss: 0.1234 - val_accuracy: 0.8672\n",
      "Epoch 25/30\n",
      "312/312 [==============================] - 11s 34ms/step - loss: 0.2248 - accuracy: 0.9029 - val_loss: 0.2383 - val_accuracy: 0.8798\n",
      "Epoch 26/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2300 - accuracy: 0.9040 - val_loss: 0.8267 - val_accuracy: 0.8815\n",
      "Epoch 27/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2218 - accuracy: 0.9101 - val_loss: 0.2709 - val_accuracy: 0.8824\n",
      "Epoch 28/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2022 - accuracy: 0.9205 - val_loss: 0.1224 - val_accuracy: 0.8555\n",
      "Epoch 29/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.2017 - accuracy: 0.9179 - val_loss: 0.2402 - val_accuracy: 0.8555\n",
      "Epoch 30/30\n",
      "312/312 [==============================] - 10s 33ms/step - loss: 0.1860 - accuracy: 0.9234 - val_loss: 0.1049 - val_accuracy: 0.8971\n"
     ]
    }
   ],
   "source": [
    "# Model Training\n",
    "model.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "\n",
    "history = model.fit_generator(training_set,\n",
    "                    steps_per_epoch = 10000 // 32,\n",
    "                    epochs = 30,\n",
    "                    validation_data = validation_set,\n",
    "                    validation_steps = 1000 // 32)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 545
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 2853,
     "status": "ok",
     "timestamp": 1591294055048,
     "user": {
      "displayName": "Siddhartha Dutta",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiT79p-YwUAMLJLPMEjadbhQl383sQDEQ2dCLM=s64",
      "userId": "06895572418557294109"
     },
     "user_tz": -330
    },
    "id": "F4l8D-9iT6se",
    "outputId": "7f8815e2-aa8c-4dc9-8842-bb791d1df6be"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Model Evaluation\n",
    "acc = history.history['accuracy']\n",
    "val_acc = history.history['val_accuracy']\n",
    "\n",
    "loss=history.history['loss']\n",
    "val_loss=history.history['val_loss']\n",
    "\n",
    "epochs_range = range(30)\n",
    "\n",
    "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
    "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
    "plt.legend()\n",
    "plt.title('Training and Validation Accuracy')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "plt.plot(epochs_range, loss, label='Training Loss')\n",
    "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
    "plt.legend()\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "authorship_tag": "ABX9TyMbc9BHBDla1RQQ/fm949D8",
   "collapsed_sections": [],
   "name": "Breast Cancer Classification.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}