Malaria Dataset (Small dataset -Random 500 samples).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from keras.preprocessing import image \n",
    "from keras.applications import vgg16\n",
    "from keras.layers import Dense, Dropout, Flatten\n",
    "from keras import callbacks \n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import datasets, layers, models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size =  20\n",
    "IMG_ROW, IMG_COL = 64, 64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "import random \n",
    "\n",
    "images = []\n",
    "labels = []\n",
    "\n",
    "images_p = random.sample(glob.glob('cell_images/Parasitized/*.png'), 250)\n",
    "\n",
    "for i in images_p:\n",
    "    img = image.load_img(i, target_size = (IMG_ROW, IMG_COL))\n",
    "    image_array = image.img_to_array(img)\n",
    "    images.append (image_array)\n",
    "    labels.append(1)\n",
    "    \n",
    "images_u = random.sample(glob.glob('cell_images/Uninfected/*.png'), 250)\n",
    "\n",
    "for i in images_u:\n",
    "    img = image.load_img(i, target_size = (IMG_ROW, IMG_COL))\n",
    "    image_array = image.img_to_array(img)\n",
    "    images.append (image_array)\n",
    "    labels.append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "500"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len (images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]],\n",
       "\n",
       "       [[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]],\n",
       "\n",
       "       [[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]],\n",
       "\n",
       "       [[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]],\n",
       "\n",
       "       [[0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        ...,\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.],\n",
       "        [0., 0., 0.]]], dtype=float32)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "images[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x24e180c2cd0>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD7CAYAAACscuKmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAApmklEQVR4nO2df5BkV3Xfv+f1r5md3ZW0IClbEkFAFGxKMcKlCDlKJbJkOQIU+IsEKjiKUWpjG7ug4hRIkDixC1xyURYEh5AsRkEpEzuUMZEiO9jKghxTRWFEEEaykEWIEIvWWlAkrXZ3pn+e/DG9877ndN+7PTP9Y2bf+VRt7et+7913+nXfeefcc+73iqoiCIJzn2LRBgRBMB+iswdBRYjOHgQVITp7EFSE6OxBUBGiswdBRdhWZxeRm0TkMRH5lojcNi2jgiCYPrLVPLuI1AD8BYAbARwF8BUAb1XVP5+eeUEQTIv6Ns69GsC3VPXbACAivwPgTQCSnV1EooInMPyNy1+2sT3y3JGxmyOvvvH4t6dt1q5GVWXc+9vp7JcA+C69PgrgtdtoL6gg933kVze2ezow+5SeDfVajXeY4y77e2+ZjXHnGNvp7OP+eow8uUXkEIBD27hOEARTYDud/SiAl9DrSwE85Q9S1cMADgPhxgej8NO8cM8Koad5Tcuf6upz7dkbdg6yndH4rwC4XEReJiJNAG8BcO90zAqCYNps+cmuqj0R+XkAfwigBuAuVX1kapYFQTBVtuPGQ1X/AMAfTMmWIAhmyJbz7Fu6WMTsu5qv/effTO5rNGvmdY3GbwcFbbth3YICyQI9s2//BQc2ttvPdTe2n3z4e+a4E995vtz+wXNm32VXvHRj+xV/6/KN7RfwrDnur7/hp3CukEq9RblsEFSE6OxBUBG2FbMHu5cvf/wTG9t1sX/zObQbDMrUWL/tng218vWgbz1H48bXyvZExnqYAIBC7M+x3Si3TzxTuvFPffW4Oe7803s3tpdP7TP7mt3lje1Of63cUe+b4x7//U+WNrpu8dfe8LakzbuJeLIHQUWIzh4EFSE6exBUhIjZK0pvrYyj+2InoNi4unwe2CgXAMrzfEqNY3atl8eNxOyUjRWXMeqsvVBuP9sp7XCPqEe/V856u+ivvNjs67fK7YI+S6Hup9+nz1I7NzPE8WQPgooQnT0IKkJU0FWEL37kE+Y1p9S8a536TRTu0cDH+TYEzXK7UDrHzVkvyuBgXfyopNcvz9vTLP3x1f93yhzXPV1W3u29wKbe9l9c2lFbLmfLDeq2Wq9Wa9C2tcPcDxdqvOL1/wg7jaigC4KKE509CCpCjMafYzxw5+Gx7/d6diydXdV+3+5jl7zXK93dwvvx5hz7WqU8r67sxrtr9agKz7XZqJeu+9pa6YIvH1gxx7GrPhDb/lpntbx2u3T/W3ub5ji0SCij5kISepnwkHcF8WQPgooQnT0IKkJ09iCoCBGzn2OoiY85ZWT/rve6XaTg8zi2d1kzl6JzYpFF2X6XwuiRFJ3Jxtp0WK/vo/h1OH4HgE6nvJaP2Tn+bjTLWL+zZo9rDMo2vPClbXD3Ph93r+VBEGyK6OxBUBGigm6BfP5Oq+mWqmrj9z3+++PzzL6ESzyufeO6Uxu51FvOjhypa/l9bGOtbts2dtXS1YB1ClrrDSe2QUIZzZbbR9druPNY9+MVN/9j7ASigi4IKk509iCoCNHZg6AiROpth8Ixqo+pi6L82kZmkZmwl+s80zF0v29jZW5z0thbnGjlIHHeaGyfepE+r99z+2hWnW/BnEc2FoWL++tUtjtwNlKWTv14wS6qnj3rk11E7hKR4yLyML13QETuF5HHh/9fMFszgyDYLpO48Z8EcJN77zYAR1T1cgBHhq+DINjBTJR6E5HLANynqlcMXz8G4DpVPSYiBwE8oKqvnKCdyqXePv9BOwtNi3Sqicm58aa9XOqNhOFYQMKfxzPbAJvKyglU5Eil6UYr6CS5L3XtbBsDWxlXp3ybks9db1r7anVK0TV8aq/cbi1ZYYsr/uHOWzZq2qm3i1X12LDhYwAu2qphQRDMh5kP0InIIQCHZn2dIAjybLWzPy0iB8mNP546UFUPAzgMnLtu/AO/bivh2FXvdDpmn+ioIPPGeeR2e7EJxrrI1mPrdseLTXiZZjNHxulAK9hlpjZGxrrT+EkzZ+hnqvVqbpJJUYwPNbxGHItveBd/QEP3BU0GKnrOqaV7MOhZG+s0Ui+SDql2Olt14+8FcMtw+xYA90zHnCAIZsUkqbffBvAlAK8UkaMiciuAOwDcKCKPA7hx+DoIgh3MWd14VX1rYtcNU7YlCIIZEhV0U8DrnbcpTvepsX6PZ2HZ29+n+JL3eUFITsX59lNps9EqvGLstmfSWZE+fcfntVotf/gG/Nl8Gxyb89hBTiBzJGY31XC0/PTIPSV7l+z38tpD/yRh/e4iauODoCJEZw+CihBufAafUmN3t0fpmZH0mrDognUXeQJGN6MDl9vHE1e8C87uqHXBcyu1WlLa6D4dyKGGD2W4+Xa7/Cw+dOFsrE+p2Wun9es5ZefTiGaVWLbRiXkotdFeczNtzhHiyR4EFSE6exBUhOjsQVARKi84+cd32qWMe73JhBlza35xC7k00Uhajo5tNMp0lU9J5VJvOQHHlB2+fYbbaDQaZh+f52Px1D6fAsylGE25L9nrM4X2nvrvhYQ4qPy2LunnXKNhxw7Y5taKXSOOdelfeygEJ4Mg2AFEZw+CilD51Jt3YTk94/cZF59mUHn3FmYWlnUJ7fJMcPvGu/i+/ZMnT25se/c5JXThXWQ72yzttubaYPfZ72O7eDtXaedTaintfO+lcjpTNR2SNEgcvtO36VLvujNmeeuOq/JrNvzhO5Z4sgdBRYjOHgQVoZKj8f/rQ3dtbHsZ5Q65ad417bTLfXzflpaWzHG9CaWYJ733/rjUMlGbaSMHf+7UiLgnt4/b8yFD7jy+Nn9mH9bkNPpSFXr9ga1Q5H05l96HGg0aje8OytDAT/5ptMpQ5qqffluy/WkQo/FBUHGiswdBRYjOHgQV4ZyN2f/k391lXg8oNu92aSaUmyXF8aWfzbZ2qr2x3aPZbD4+q7fKKqtcxZif2TapXjufNzqLbHz6blTUYTLhxNzS0bm0mYmP6Z7mxDY8qeq6XKowPwaA5HH1Bo9NTKbn723Jtd9olp+lcEtIsaDl33z79nXoI2YPgooTnT0IKsI5W0Hn3UWufmOBB+/Gs9aE052w7qKpdktfOzfZJUeuja2k83IueK693KSbSW1iV3fS8CF3bO5++LDAhhDjbRolXcnHE1+8LaocCjg76PfndDMm/k1sl3iyB0FFiM4eBBUhOnsQVIRzNmaHplNNHLv5JdXWTp8ee8766/RMNHNp1muHiy9pn4/VUqksH19y6s3Hr6m4Opfi8vE2H5uLJ3O69Kkxh1yq0Kc60+IV9rNo5p7aeJsFMs1h0NTidK7NwcDOeuPfQU4Dnz+3T73x7+qPf4PEVDLDCn/3525N70wwyfJPLxGRL4jIoyLyiIi8c/j+ARG5X0QeH/5/waavHgTB3JjEje8B+EVV/WEA1wB4h4i8CsBtAI6o6uUAjgxfB0GwQ5lkrbdjAI4Nt18QkUcBXALgTQCuGx52N4AHALxnJlZugW7XunM8u42XYPKz3visvnOJ6yRUUJBrzRVzgBN16LTtPnLhmk17nnUXS7cy54J7Uss6eTc7v2TSeJfWH+fTS5bxrvtoG+XrXGhkWs6ELqOz6jjU6NG2bdOcV+RSe14shO2gJopMGNl198CYnE4L+zY3y6YG6ETkMgCvAfBlABcP/xCc+YNw0bYsCYJgpkw8QCciewF8BsC7VPXEpPOoReQQgENbMy8Igmkx0ZNdRBpY7+ifUtXfG779tIgcHO4/COD4uHNV9bCqXqWqV03D4CAItsZZn+yy/gj/BIBHVfVO2nUvgFsA3DH8/56ZWOj4woc+ntzH8ZPXfxeMn53U69l0j4+jmUlnCOZmYXF6yaehOG7MiTQyW53NNml6jdnMjLXU7LvcPcxpz+dm3+WEL9nEnL1mjGQkLZdbW2/8dz0qipkeE0jdEp8OzK1VMAmTuPHXAvgpAN8QkYeG770X65380yJyK4AnAbx5W5YEQTBTJhmN/yKA1J+UG6ZrThAEs2LHVNB96aOl2IT3RHmp4E4vNwsrvXQvH5pzs7nKysPnsTu+1dTYVgQk/Ouca51zfdndndR9ztrlbeQUUsYOfu3FPFLXHrkftNzy6LJOk82cYzt81aM5LjPjzrTp04P8IjObktOPxYR6/pMStfFBUBGiswdBRdgxbjz/3Rk41Qh+mVtlFWbU12nLkQtUK7giKid24Nw5rvaiUfvCuZXsBm9GvMK4khmXLVUlB1iXvN0uq/dyo+C5paFM5Zf3qjOTU0AjybnPnBO24DZZ86/uQi3+bL2+W56plnCz/ZC7+U340G68np63MTdqn8uo8HfGbXQ723PbPfFkD4KKEJ09CCpCdPYgqAhzjdl/6K++FJ+8/V8CADptG490VnlGkt3XpTis0Shj5V7XpXFMxZFLvVHM15+wEmk09cPxVJl68/FYg2ZG+RRgjUUpvCihmZlHn3nJVvXVSON80LOxOKevuFBrMKGopKefEXUAzyjTfmpXMq711/YpQP6aisxzyczgc79otQeOPcfb6LHjP2nhDD7O39J6ndcS8EuBj4/nc9V6WyGe7EFQEaKzB0FFWEDqbfj3ZWDdudXV1Y3tEf2uZrm8UmdtbWNbnEgXp0jykwZYGMLegm43vWQz28WVWq3Cii70OuVnaxbWBe+2yzbba9Yt6/VKF7y1VLYpav8md9bKlNrSHrtctKlW46WJnPvMlVr+fnMFWbMYnxYa99qQElpQfxhVvzXtfUxNoMlPhOkl9+Um/6TSXzk7hlcYa5c/zv42kSQn9JHXuj878WQPgooQnT0IKkJ09iCoCHON2VXLdFm73TnL0SVGWIBmOPUzpbO5LAWnuDhOBgCl5Zx9fCkUDiovAe3LQftlrPXs88/ZXWuUXjttz+OUXadOM/Naa+a4PS8q4/SuE0JAjYys8cyztPZ8UXcijYmy3c2sMWf12ieLNScV6cjF1D5dlUr75VKAOTs8udLllI35dffYLmvj627bnrpbPNmDoCJEZw+CijD31NuGOzaiq90fPWaIEXyg90eFG8qP0+879zzhwhV+maic7cqVazSry81OkgGLKVgXtk/uc7/j3EVqpk8Vaa29Nn3XWS7bWNpr01U1ugddClG0nnazc4ISqXP869x3NqkLniO3xHQ+vBgfhmxGuy8XJuT07xgOZXKVcZPej60QT/YgqAjR2YOgIszdjT/jpnhXprlUVsn5JZk6VBXFQhY5gQA/AnyaVmflZYvWVq27P6DJNUs1V9HVock0NGo/cMv5gNqQnq/yK7fdvA/IoHzj1HMny/ZdtWGjsXdje7VhsxrSLG2srfCIuG2jWStH9Dsjgg/jR89HV09Nh161WmoUfJA8zrfBbrGtfkuHJHY02x6byyzwvlarZfaxHV4nLzUC7yf1cPtveO/PjD1n1sSTPQgqQnT2IKgI0dmDoCIsTHDSxzqdLseN6SooQTqNw/RchR5Xp3HcVfOpFBIo9Ms+czjIMVjfiWj02hyz2/ZZRLBQGxvzx+GKt6bXsmf9+lU3Q5C+0voypYWcyAWLURYN274kZ4elZ5uNxq48szCnPZ+O57lJW5EHdxynv9I2TroMla+g48/ml5VOLa3t03CzTKlNylmf7CKyJCJ/KiJfF5FHROSXh+8fEJH7ReTx4f8XzN7cIAi2yiRufBvA9ar6agBXArhJRK4BcBuAI6p6OYAjw9dBEOxQJlnrTQGcyQM1hv8UwJsAXDd8/24ADwB4z1na2tB9z4kH+NRKwauzsjvkCpY4FeTTVexG1UiorOvSTpx6a9R8+mR8KshrxLUHXTrH2tFqlW5gt52u2lraR+kf57YOaqSB7yZLqAkvKPxpZJY78mIhtOJos8miDukKMZ/yUvC9otVpa+kKNA+HW+yeNxo2NZZLqfHvLFfJl7Mjt5ZAyo3Ptf/f3//vzT4+b5ZpuUnXZ68NV3A9DuB+Vf0ygItV9RgADP+/aGZWBkGwbSbq7KraV9UrAVwK4GoRuWLSC4jIIRF5UEQefP7UybOfEATBTNhU6k1Vn8O6u34TgKdF5CAADP8/njjnsKpepapXnbeyd9whQRDMgbPG7CJyIYCuqj4nIssAfgLArwG4F8AtAO4Y/n/PBG1txCc+fcJlsFl9bDNDyJc8UmrFpWC4SZ6xVvMBsYwvrwSAJq3v1q9RCeWqjWVbeyguhyvHJbuaTjSiR6m+5dZy2UbPilcUeygetnqTECqL7dNgRx2ZWFl8fMkpKrpWRoNiJL1WGy+c6GPjXDrMCkOU73c69n7w96Ka/py5mXO5uJ/t8OWyviw2Bcf9vhx5M2IZ22ESSw8CuFvWE78FgE+r6n0i8iUAnxaRWwE8CeDNM7QzCIJtMslo/J8BeM2Y958BcMMsjAqCYPosrILOu0pdWhKn73JqA1NZRtVSXnShx66Y01UzKTtyxUYq6NKzsHqc1qJqr1rLzWyjWXUDpxuv5LIVzuVkoQtV1qi3djRJsKKwWSjzceqUNiuWbBt1qgTzbuVA2a1k4Ya0C16vu6WsqIKuVssNDWUEJSgk6bXpfjTctczsOPuTNsIn/P251G+uGrBDS32N/CYSLnhOdy9mvQVBMFOiswdBRVicBp2r/OLllEZcJXIr+4mKpfU3ys1u1+u7jZc27vWsO5dzv5RcUx4BHtEUo2vXnMtZo1Vo4VZ4FbMqKrnBbvJFbYlWgl2y9jaa5WuOZHpiPyePrHd7bbPPV8NtXMuJebBbP+Lim/Nyk5fGu9megleutQPiJpkwqbgEf3+AddW9a87H5trP6frlZKbnxeItCIJgLkRnD4KKEJ09CCrCfJd/AtD3ayoNsTGOj2UpwNT0cdxEIfaj9SjQ42WCOZ4E0jEYYCfZmQI0XxVG8euS03zvkmjlaKxcbtcpPG66uFzpo9WbruqM0lUDoQo6l/4aUKqz0fDjFDRDkMYwRj4n3xFxMWpiDGY0Zh+/1JR/XWTK93pGGz5dhceCHTn8uE0u9cYpvNxyy7tCvCIIgnOD6OxBUBHm6sYLSvemO8ilYFwF3WD8BIacZ+SlCFg7jFMrhXNve/10am9lZaW0iQQ21KeMuDLO6T0MyEWuu5Qau+ANct2l7lZ7pfSa148zEzoKDlfSbnBO8z0n6sDpwb77MuqJ621mAkoqleWPM+7zyO+qPNbrx6Wu5dNrqeNyNvqQhO//7//qf0i2v3DxiiAIdj/R2YOgIkRnD4KKsIBZb+uxDKd+PDkRA/A6bWs+dUWCfz7WpDYbRRm7iStnZTGImkvfddfKWO7Zo8+V57iy14JiyOaKE0ck4Ue/FDNqlOJp0ppzS8vmME41+aWvOUU1kJxA5Pjy4fV95bYR6qy7WJlm9PkYVQoe00iakRSoAIAurafX75DWf81rt5fb/ZFrlW9MKhIx8lnohvgyW24zV+5rYvv0MMhMiSd7EFSE6OxBUBEWNuvNa3d1e+l0h9EOo/dHZhKxf5SpYOLzOq6KbblRusziEngsoGA85IFLoaF8rR0vjsHhhJuxtocEOWvsHprDUKunKgqtu9ioZ9zsTEotpRnnJP/MYlC5sMmEVy5Ey2nQsYucn1FWT+7jNFq+ko/bS+vYTbqsU76CLp2+myXxZA+CihCdPQgqwsI06EaXx2GX1o6aGtcp4/Z5t5sx1U08+cK5wQNa7bTbsS7+c3/5/MZ2fY1c5FP2WqfWTm9s732R1covzqPR7b1uVLlNLidpxonTuCto8o6vwmOMDLRzTfnejY4i87EU/vjbS6Pzo/p0bO/49wGgxyPuzkXm1Wt7Zukpe63V1dWNbZ9ZSFVO+pH5XIUh3x9fXcfnmcyFa2/ScGWWxJM9CCpCdPYgqAjR2YOgIsg8J9VLSskQwOc+UC5j2+vZGJJjucH48H24U8YeBwB9iotqXEFX2EZqFA+ePrlq9g2eLY/tfJ/i4dN26GPtdFkJt/f8FbOvtr889vxL9tv295Tn1fdSLNuwMV1jiVJ7kq7aKviz+Yl5JhVZT+7jE2tO5KLZomeFC3m5zVy82ucxEj/bjL5DXrJ5ZJLhIJ1SS9kxsm4BXTvXJ3Kz9mw1YHqM5Ob3/Wyy/Wmgfl20MzZN2sBw2eavich9w9cHROR+EXl8+P8F0zI2CILpsxk3/p0AHqXXtwE4oqqXAzgyfB0EwQ5lIjdeRC4FcDeADwD456p6s4g8BuA6VT02XLL5AVV95VnamShmuPdffyS5j93zet2lrsgl9OmMfpf02Kh6r9frmOPqNPllpWld8O984+jGduNUue/571l3/9TzZertwIUvNvuUKuMOXGLTcrK/tHHlwvKz1fe6CSh7ye0u0hVdHaXJIy5dZUIj5xZbHT6auLPHTuppLpMdzo1PpaRyFW5+OawefWc5N571Bn377NHmBComFdHwpPZ5O974r96RbGPabNeN/zCAd8NGfher6rFh48cAXLQdA4MgmC1n7ewicjOA46r61a1cQEQOiciDIvLgVs4PgmA6TFJBdy2AN4rI6wEsAdgvIr8F4GkROUhu/PFxJ6vqYQCHgcnd+CAIps8k67PfDuB2ABCR6wD8C1V9m4h8EMAtAO4Y/n/P1KxSP+uIhRPTp/G6auKWSm40y319sxabvQVCweeJ1ZNm374DZYx94kRZI7t3/x5z3NrJMh7srbkAs1mGUz948nmza2lPGacPTpTntV5sQ7B9l5Y29+u2fS4rXV4pZ/ANnPKl0D3uuLLgTptENGiGXbNuY3ZtULmsSwFqMX5MIDdrbHQZ5fGO58jaeiae99rz46+11bjcj3GlUm/zmsm2GbZTVHMHgBtF5HEANw5fB0GwQ9nURBhVfQDAA8PtZwDcMH2TgiCYBQub9Zbjjb/y88l99/7SRze2czOLPKZijE4Tl6WoccgAm9prnkehwYXlcc8fXbNtkKvuNeUHHUplNZw+3any2JPd0rU+/bxNE60s7Svb32ftXybNu26bwxVzmHE5u12nS0+iF7VmblYdp9TsPvbqlZbFFvUVi5Si80t198eLV/hltm1VZXo9gtzvY9JZb7nlnDkMOdfc+CAIdhHR2YOgIuxINz7H0lKp9exHXtnF8h4bL2lkPbH0hAVfFdYtylFqpTksjQvsgXs65Sh474S1kSvBTqzayrv9e0v3vE8VXnUXapw4XmYC9tTsV7jaJxuXyw/aXLLVegNyhVuuEvH06TIsadJySs7bx0qjzEI0nR1qvot0BR0Xe/nJSyxSwfdtdCS93PaTelKrrG5GC49feynpFD5jcN8HPraxPeuJMCniyR4EFSE6exBUhOjsQVARdl3MntIBB1z6JJNmYXn5mqvSKihO9Nr2a2tlPLxC1XSd50+Y4xok6tAXO6uuRVVorcKm3li0o9MpY949+88zx3VO0mywZ0+bfcvnl2MadZqVtrZm04Mcs/vlmWos9MjLW7t42C7DlL7ffV7e2h3GQpU1t3x2Mk536yeZ09z3mVpGOaej739XHH/7ZZ87HRojSaxNMO56iyCe7EFQEaKzB0FF2HVufEE6aH6iBLtRdTfBrtvpjz3OO599fsflgpZXShd59cQLG9t7L14yx7VJoOJU14rKd06VYUjTufHs+tXYDfTCEEvl11a4AjfO0vFnq/sDyX3uwIYaXMlWo1Cm7iYNZfXOaZkrDq/8ZBcWnvBLgLEbz6vTamb5JBYwAYA6n0enedtzbjzb7EUvrLdOK+g6t/0N7/0ZLJp4sgdBRYjOHgQVITp7EFSE3RezFzyLKX3cyAwnmoZlUnR9X846UrNJbdI+yveIE5DQRnnc3hc5wYcVEkdctdeq8/porbIUtbXfpbzOI9GIfe4rbHKcOz6FBtjZeLnyUINbz43j0r4vTyZxDB764BjdX2tiAQm3z8bHaT14Tpv5WW6DzP1I2euv59e722nEkz0IKkJ09iCoCLvOjWedd1/RJTRDqzOSIqGleVhYwbmO9eZ4vXMAEPrbuLxCenFtl+65oLz2Hif+0D1Ny0Z13fLClCZq8nLOy849XC6vt7TPzsJiYY5Wq7w/q6u2gq5H4UrN30cKSwrSoPMVhXzr+v10Kovvfb+friTzohSmIo3dc5fWap/u0S4/E7Jss41SEGTPPjsL0IpcWDtyqcPdRDzZg6AiRGcPgoqw69x4Fq/ouyFg6+KnR45ttZRt3078SP8tFGq/aFg7Vs4vxStW61agQpo8+uxWHCXPvVgiwYSW021rll+bX4WWP9tqu3RbRyd+8Gi53ZeSRPYubKNWhhCFc/H51vG1fSjAo+V+hJzlrnukp8dLeQFAt00hmwsF1lbLe9DVso2VFbu0l/JsIPe15zM081uddbvEkz0IKkJ09iCoCNHZg6Ai7LqYvW6WeLKxFOkIjMR/HM9z3J+b/TSSeuP4VTnd43TMKUZt7bWpMaGXOZEEntA38JVZlGKsZcYcTFWYi8sHrHHuRCM43caPgyKz9JYXlOAFf23qzd4ro7XuqxnX6HOShn/f6nVAOjT+sGbb762SCMi+spqx17Yz/cDLg8FBH81/Z43sTdlZTNTZReQJAC9g/T70VPUqETkA4L8CuAzAEwD+gao+OxszgyDYLptx439cVa9U1auGr28DcERVLwdwZPg6CIIdynbc+DcBuG64fTfW14B7zzbtOSucrhrVoOPX9u8Yp6tYN2y0/bSIQWqChF8JtlVkxBoy6byUHaOCCayB7ya4JDTavdYeCznUMjbZz5wWdfD3pk3LV2X13ajNQc/ZQSk21tDrtr27b+8PwyHb8nK57VOF/YJ/V8nmTJUmALzuvYfSB+8wJn2yK4A/EpGvisiZT3exqh4DgOH/F83CwCAIpsOkT/ZrVfUpEbkIwP0i8s1JLzD847B7/vwFwTnKRE92VX1q+P9xAJ8FcDWAp0XkIAAM/z+eOPewql5FsX4QBAvgrE92EVkBUKjqC8PtnwTwKwDuBXALgDuG/98zS0PPoMIiAzZebS2V6ZlO2+6r1yk9k4tRe+k41AtcbrTnZ1rRdt+l5Xy5qLl2YkzAx7kcw+fSg8ZGsTZm7TBrp1E6cBNLZA/oOVJoWhiiNqDxh1Ube3dPUokspeG4jBYAmqTFr+5zFi0W5yThEPed3fy+nxvzKc4tJnHjLwbw2eGPqA7gv6jq50TkKwA+LSK3AngSwJtnZ2YQBNvlrJ1dVb8N4NVj3n8GwA2zMCoIgumz+yro6qwbb/etrpYptWbLVq759NUZeMklAEZnbcRLpQovXiIpp+HmU205IYQ+hRDsjWaXuXJMGgow3qWvk+AGVwd615cL+/oupKqx2H1GK5Bdcq9tzxLwOijbb7hUJyht1nO68WaJZTPLMGPUOUrUxgdBRYjOHgQVITp7EFSEXRezX/uztyT3feHDhze2+66M1JSYdst9I2oxGV36Pv1tlMxMKG6D15hb3ze+nBWwuuY5dZRaZhaZUPkpx/ai1kae6TaAKz9l8UgeHnD3I7eWXKdblreyskzdpS/ZrkL9fSy3W0vpn2pBP2Np2fGMPt9jLnX10wUrQDzZg6AiRGcPgooguSqoqV/M+8wz5Miv/0d37dK963SoMquXTpuNptQ410RLH2W00H3qLTfrzS6FNP5934YPBRg728y5z7mZXZTe5JmE/loDpNNynM4rZPx9AwDtUpqyb21snyQBUV4Ouet03SmEUCfAWaMKOiU9/Ne9f/FLKM8KVR377caTPQgqQnT2IKgIu240flJqdft3jN1143K6yGLglyMluKqNj/IVbex2j64Wmh5lZ9eXR9lzFXOe1Kqo3m3nfV4+jl/mdPTZW8xV6BlBDTekX+OsgHPxhYvwaFfbVUPuaVEmwN0qNYpy1RuBZ+LJHgQVITp7EFSE6OxBUBHO2dRbjiMf/NjGds9NhvPrxxmUq+s4RWcPy6fGeIaWmymWiM1H4v5MWi71fWbXrXPxdupr6ms6vTb6WcYv2TxSbUivfXVdh4QktUP32w179HjtNyde8fd/7RdQNSL1FgQVJzp7EFSEczb1lqNOggYyImKQ1qDjSS3swna71oXNCSOwS5tL2YFSUt1Mui5XoWcn9aS16rzTV2+Md89r8Etq+c9N7Seu5T8zr57k06XNWplSK/bQZ3G3Y2BULqqdXssRT/YgqAjR2YOgIkRnD4KKUMnUG8NpOABYI9FKESvIgP74dFLPxdQco/bUx7UksOiEHs0Sy0a00raQE7Q0V+KY3YWyuTZSy1bn4n6fUqvXOE4ff9/W2xw/DgIAhTkPSa75Z2lBkyoSqbcgqDjR2YOgIlTejc/xP97/G+Z1o0HLDJGn5AUwOEPlK+h4NlvOBefzRo4rxrvZgHWnJxXKGHWtJ/ua+LQRbfv6eDfew7Peipqv0OML8Gw+a9/Vb791InurwrbceBE5X0R+V0S+KSKPisiPicgBEblfRB4f/n/BdE0OgmCaTOrG/1sAn1PVH8L6UlCPArgNwBFVvRzAkeHrIAh2KGd140VkP4CvA3i50sEi8hiA61T12HDJ5gdU9ZVnaWtXufGf/5AdqR+QXprSbBQtXCUc7RuRei7SE2FSE0aKEQGMSXXn0u4zt+Gr2nipLHbHc9V6o3La5bZdssuN2rPuRM2HAuU2/3QaTT9hiDMG9n7/yFvfjqqxHTf+5QC+D+A/icjXROQ3h0s3X6yqx4aNHwNw0dSsDYJg6kzS2esAfhTAx1T1NQBOYRMuu4gcEpEHReTBLdoYBMEUmKSzHwVwVFW/PHz9u1jv/E8P3XcM/z8+7mRVPayqV6nqVdMwOAiCrTFR6k1E/gTAP1XVx0Tk3wBYGe56RlXvEJHbABxQ1XefpZ1dFbPn+OJHPr6x3XdChn1eBtrF87wEkRfKqCXEIEZidBJoyMXludlmubjf/iZo6SZX4ZYbHyhqZRt8mreD9SqaTbe8NZ3HsXgV4/DNkIrZJ53i+gsAPiUiTQDfBvDTWPcKPi0itwJ4EsCbp2FoEASzYaLOrqoPARjnht8wVWuCIJgZUUG3C/mfd5ZLW+W/PxbKsJ4dnzc6IWe8gMf17zq0WVODBRATYYKg4kRnD4KKEJ09CCpCJQUndzuc5RqZbUaprcGAZ4rZNjhOHy1h5X0xzHKuEE/2IKgI0dmDoCLMO/X2fQDfAfBiAD+Y24XThB2WsMOyE+zYrA0vVdULx+2Ya2ffuKjIgzuhVj7sCDt2uh3TtCHc+CCoCNHZg6AiLKqzH17QdT1hhyXssOwEO6Zmw0Ji9iAI5k+48UFQEeba2UXkJhF5TES+NRS8mNd17xKR4yLyML03dylsEXmJiHxhKMf9iIi8cxG2iMiSiPypiHx9aMcvL8IOsqc21De8b1F2iMgTIvINEXnojITaguyYmWz73Dq7iNQAfBTA6wC8CsBbReRVc7r8JwHc5N5bhBR2D8AvquoPA7gGwDuG92DetrQBXK+qrwZwJYCbROSaBdhxhndiXZ78DIuy48dV9UpKdS3CjtnJtqvqXP4B+DEAf0ivbwdw+xyvfxmAh+n1YwAODrcPAnhsXraQDfcAuHGRtgDYA+B/A3jtIuwAcOnwB3w9gPsW9d0AeALAi917c7UDwH4A/xfDsbRp2zFPN/4SAN+l10eH7y2KhUphi8hlAF4D4MuLsGXoOj+EdaHQ+3VdUHQR9+TDAN4NFrtbjB0K4I9E5KsickalY952zFS2fZ6dfZx6RiVTASKyF8BnALxLVU8swgZV7avqlVh/sl4tIlfM2wYRuRnAcVX96ryvPYZrVfVHsR5mvkNE/s4CbNiWbPvZmGdnPwrgJfT6UgBPzfH6nomksKeNrC/6/hkAn1LV31ukLQCgqs8BeADrYxrztuNaAG8UkScA/A6A60XktxZgB1T1qeH/xwF8FsDVC7BjW7LtZ2Oenf0rAC4XkZcNVWrfAuDeOV7fcy+AW4bbt2A9fp4psj5x/BMAHlXVOxdli4hcKCLnD7eXAfwEgG/O2w5VvV1VL1XVy7D+e/i8qr5t3naIyIqI7DuzDeAnATw8bztU9S8BfFdEziyjdgOAP5+aHbMe+HADDa8H8BcA/g+A983xur8N4BiALtb/et4K4EVYHxh6fPj/gTnY8bexHrr8GYCHhv9eP29bAPwIgK8N7XgYwC8N35/7PSGbrkM5QDfv+/FyrK9n+HUAj5z5bS7oN3IlgAeH381/A3DBtOyICrogqAhRQRcEFSE6exBUhOjsQVARorMHQUWIzh4EFSE6exBUhOjsQVARorMHQUX4//G/ifhS6y0FAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(images[0]/255.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle and split train/test set \n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# Shuffles\n",
    "images, labels = shuffle(images, labels)\n",
    "\n",
    "# train-test-split\n",
    "X = np.array (images)\n",
    "y = np.array (labels)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split (X,y, random_state = 42, test_size = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(400, 64, 64, 3)\n",
      "(400,)\n",
      "(100, 64, 64, 3)\n",
      "(100,)\n"
     ]
    }
   ],
   "source": [
    "print (X_train.shape)\n",
    "print (y_train.shape)\n",
    "print (X_test.shape)\n",
    "print (y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build  CNN models "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale for X // no need to encode y as Binary categorical\n",
    "X_train = X_train / 255.0\n",
    "X_test = X_test / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "20/20 [==============================] - 3s 128ms/step - loss: 0.6858 - accuracy: 0.5500 - val_loss: 0.6856 - val_accuracy: 0.6000\n",
      "Epoch 2/20\n",
      "20/20 [==============================] - 2s 110ms/step - loss: 0.7000 - accuracy: 0.5225 - val_loss: 0.6929 - val_accuracy: 0.4800\n",
      "Epoch 3/20\n",
      "20/20 [==============================] - 2s 105ms/step - loss: 0.6794 - accuracy: 0.5575 - val_loss: 0.6801 - val_accuracy: 0.5100\n",
      "Epoch 4/20\n",
      "20/20 [==============================] - 2s 99ms/step - loss: 0.6636 - accuracy: 0.6050 - val_loss: 0.7183 - val_accuracy: 0.5900\n",
      "Epoch 5/20\n",
      "20/20 [==============================] - 2s 103ms/step - loss: 0.6205 - accuracy: 0.6575 - val_loss: 0.6934 - val_accuracy: 0.5900\n",
      "Epoch 6/20\n",
      "20/20 [==============================] - 2s 104ms/step - loss: 0.6102 - accuracy: 0.6275 - val_loss: 0.6629 - val_accuracy: 0.6100\n",
      "Epoch 7/20\n",
      "20/20 [==============================] - 2s 105ms/step - loss: 0.5335 - accuracy: 0.7100 - val_loss: 0.7005 - val_accuracy: 0.5500\n",
      "Epoch 8/20\n",
      "20/20 [==============================] - 2s 106ms/step - loss: 0.5180 - accuracy: 0.7500 - val_loss: 0.8182 - val_accuracy: 0.4700\n",
      "Epoch 9/20\n",
      "20/20 [==============================] - 2s 106ms/step - loss: 0.5021 - accuracy: 0.7475 - val_loss: 0.7785 - val_accuracy: 0.6200\n",
      "Epoch 10/20\n",
      "20/20 [==============================] - 2s 108ms/step - loss: 0.3915 - accuracy: 0.8275 - val_loss: 0.8365 - val_accuracy: 0.5700\n",
      "Epoch 11/20\n",
      "20/20 [==============================] - 2s 107ms/step - loss: 0.2917 - accuracy: 0.8725 - val_loss: 1.1055 - val_accuracy: 0.6000\n",
      "Epoch 12/20\n",
      "20/20 [==============================] - 2s 108ms/step - loss: 0.2206 - accuracy: 0.9100 - val_loss: 0.9390 - val_accuracy: 0.6300\n",
      "Epoch 13/20\n",
      "20/20 [==============================] - 2s 111ms/step - loss: 0.1719 - accuracy: 0.9500 - val_loss: 0.9463 - val_accuracy: 0.5900\n",
      "Epoch 14/20\n",
      "20/20 [==============================] - 2s 113ms/step - loss: 0.1196 - accuracy: 0.9550 - val_loss: 1.0978 - val_accuracy: 0.6600\n",
      "Epoch 15/20\n",
      "20/20 [==============================] - 2s 108ms/step - loss: 0.1189 - accuracy: 0.9575 - val_loss: 1.0893 - val_accuracy: 0.6400\n",
      "Epoch 16/20\n",
      "20/20 [==============================] - 2s 113ms/step - loss: 0.0558 - accuracy: 0.9875 - val_loss: 1.3588 - val_accuracy: 0.6300\n",
      "Epoch 17/20\n",
      "20/20 [==============================] - 2s 105ms/step - loss: 0.0444 - accuracy: 0.9800 - val_loss: 1.5160 - val_accuracy: 0.6400\n",
      "Epoch 18/20\n",
      "20/20 [==============================] - 2s 108ms/step - loss: 0.0525 - accuracy: 0.9800 - val_loss: 1.2872 - val_accuracy: 0.6700\n",
      "Epoch 19/20\n",
      "20/20 [==============================] - 2s 107ms/step - loss: 0.0563 - accuracy: 0.9825 - val_loss: 1.4104 - val_accuracy: 0.6400\n",
      "Epoch 20/20\n",
      "20/20 [==============================] - 2s 109ms/step - loss: 0.0275 - accuracy: 0.9950 - val_loss: 1.5871 - val_accuracy: 0.6200\n"
     ]
    }
   ],
   "source": [
    "model_cnn = tf.keras.models.Sequential()\n",
    "\n",
    "# Convolution and MaxPooling\n",
    "model_cnn.add(layers.Conv2D(32, (3,3), activation = 'relu', input_shape = (64,64,3)))\n",
    "model_cnn.add(layers.Conv2D(32, (3,3), activation = 'relu'))\n",
    "model_cnn.add(layers.MaxPooling2D(2,2))\n",
    "model_cnn.add(layers.Conv2D(32, (3,3), activation = 'relu'))\n",
    "model_cnn.add(layers.MaxPooling2D(2,2))\n",
    "\n",
    "# Flatten\n",
    "model_cnn.add(layers.Dropout(0.2))\n",
    "model_cnn.add(layers.Flatten())\n",
    "\n",
    "# ANN part and compile\n",
    "model_cnn.add(layers.Dense(64,activation = 'relu'))\n",
    "model_cnn.add(layers.Dense(64,activation = 'relu'))\n",
    "model_cnn.add(layers.Dense(1,activation = 'sigmoid'))\n",
    "\n",
    "model_cnn.compile(optimizer = 'adam', loss= 'binary_crossentropy', metrics = ['accuracy'])\n",
    "\n",
    "hist = model_cnn.fit(X_train, y_train, batch_size=batch_size, epochs=20, validation_data = (X_test,y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_6 (Conv2D)            (None, 62, 62, 32)        896       \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 60, 60, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 30, 30, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 28, 28, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 6272)              0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 64)                401472    \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 64)                4160      \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 425,089\n",
      "Trainable params: 425,089\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model_cnn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "epochs_list = list (range (1,len(hist.history['accuracy'])+1))\n",
    "plt.figure(figsize =(10,10))\n",
    "plt.plot(epochs_list, hist.history['accuracy'])\n",
    "plt.plot(epochs_list,hist.history['val_accuracy'])\n",
    "plt.legend (['Training accuracy', 'Accuracy'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning using ImageDataGenerate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "20/20 [==============================] - 2s 112ms/step - loss: 1.0216 - accuracy: 0.5700 - val_loss: 0.6368 - val_accuracy: 0.6900\n",
      "Epoch 2/100\n",
      "20/20 [==============================] - 2s 103ms/step - loss: 0.6462 - accuracy: 0.6450 - val_loss: 0.5635 - val_accuracy: 0.7400\n",
      "Epoch 3/100\n",
      "20/20 [==============================] - 2s 104ms/step - loss: 0.6098 - accuracy: 0.7700 - val_loss: 0.5594 - val_accuracy: 0.7100\n",
      "Epoch 4/100\n",
      "20/20 [==============================] - 2s 105ms/step - loss: 0.5991 - accuracy: 0.7375 - val_loss: 0.4417 - val_accuracy: 0.8600\n",
      "Epoch 5/100\n",
      "20/20 [==============================] - 2s 111ms/step - loss: 0.4783 - accuracy: 0.8025 - val_loss: 0.4961 - val_accuracy: 0.7700\n",
      "Epoch 6/100\n",
      "20/20 [==============================] - 2s 111ms/step - loss: 0.4657 - accuracy: 0.8325 - val_loss: 0.3305 - val_accuracy: 0.9200\n",
      "Epoch 7/100\n",
      "20/20 [==============================] - 2s 111ms/step - loss: 0.4111 - accuracy: 0.8650 - val_loss: 0.3704 - val_accuracy: 0.8600\n",
      "Epoch 8/100\n",
      "20/20 [==============================] - 2s 113ms/step - loss: 0.3352 - accuracy: 0.8825 - val_loss: 0.3338 - val_accuracy: 0.9100\n",
      "Epoch 9/100\n",
      "20/20 [==============================] - 2s 113ms/step - loss: 0.3227 - accuracy: 0.9025 - val_loss: 0.2857 - val_accuracy: 0.9300\n",
      "Epoch 10/100\n",
      "20/20 [==============================] - 2s 115ms/step - loss: 0.2835 - accuracy: 0.9050 - val_loss: 0.2854 - val_accuracy: 0.9500\n",
      "Epoch 11/100\n",
      "20/20 [==============================] - 2s 115ms/step - loss: 0.3974 - accuracy: 0.8325 - val_loss: 0.3504 - val_accuracy: 0.9200\n",
      "Epoch 12/100\n",
      "20/20 [==============================] - 2s 110ms/step - loss: 0.2904 - accuracy: 0.9025 - val_loss: 0.2922 - val_accuracy: 0.9300\n",
      "Epoch 13/100\n",
      "20/20 [==============================] - 2s 112ms/step - loss: 0.3090 - accuracy: 0.8800 - val_loss: 0.2708 - val_accuracy: 0.9400\n",
      "Epoch 14/100\n",
      "20/20 [==============================] - 2s 113ms/step - loss: 0.4505 - accuracy: 0.8675 - val_loss: 0.2713 - val_accuracy: 0.9300\n",
      "Epoch 15/100\n",
      "20/20 [==============================] - 2s 111ms/step - loss: 0.3044 - accuracy: 0.8875 - val_loss: 0.2911 - val_accuracy: 0.9200\n",
      "Epoch 16/100\n",
      "20/20 [==============================] - 2s 114ms/step - loss: 0.2402 - accuracy: 0.9175 - val_loss: 0.3102 - val_accuracy: 0.9200\n"
     ]
    }
   ],
   "source": [
    "gen = ImageDataGenerator (rotation_range=20, zoom_range=0.15, width_shift_range=0.2, height_shift_range=0.2, shear_range=0.15, horizontal_flip=True)\n",
    "\n",
    "early_stop = callbacks.EarlyStopping(monitor='val_loss', patience=3)\n",
    "\n",
    "model_cnn2 = model_cnn.fit_generator(gen.flow(X_train,y_train, batch_size=batch_size), validation_data=(X_test,y_test),steps_per_epoch=len(X_train) // 20, epochs=100,callbacks=[early_stop] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "epochs_list = list (range (1,len(model_cnn2.history['accuracy'])+1))\n",
    "plt.figure(figsize =(10,10))\n",
    "plt.plot(epochs_list, model_cnn2.history['accuracy'])\n",
    "plt.plot(epochs_list,model_cnn2.history['val_accuracy'])\n",
    "plt.legend (['Training accuracy', 'Accuracy'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Transfer learning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications import vgg16\n",
    "base_model = vgg16.VGG16(weights = 'imagenet', include_top = False, input_shape= (64,64,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"functional_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         [(None, 64, 64, 3)]       0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 64, 64, 64)        1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 64, 64, 64)        36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 32, 32, 128)       73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 32, 32, 128)       147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 16, 16, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 16, 16, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 16, 16, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 16, 16, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 8, 8, 256)         0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 8, 8, 512)         1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 4, 4, 512)         0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 2, 2, 512)         0         \n",
      "_________________________________________________________________\n",
      "flatten_4 (Flatten)          (None, 2048)              0         \n",
      "_________________________________________________________________\n",
      "dense_12 (Dense)             (None, 64)                131136    \n",
      "_________________________________________________________________\n",
      "dense_13 (Dense)             (None, 64)                4160      \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 14,850,049\n",
      "Trainable params: 14,850,049\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Input new classification prediction \n",
    "from keras.models import Model\n",
    "x= base_model.output\n",
    "x= Flatten()(x)\n",
    "x= Dense (64, activation = 'relu')(x)\n",
    "x= Dense (64, activation = 'relu')(x)\n",
    "x= Dropout(0.2)(x)\n",
    "predictions = Dense (1, activation = 'sigmoid')(x)\n",
    "\n",
    "trans_model = Model (inputs=base_model.input, outputs=predictions)\n",
    "\n",
    "print (trans_model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing Transfer Learning\n",
      "<tensorflow.python.keras.engine.input_layer.InputLayer object at 0x0000024E12D3DD30>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E14183490>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E141839A0>\n",
      "<tensorflow.python.keras.layers.pooling.MaxPooling2D object at 0x0000024E14183910>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E17EA20D0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E17EA8460>\n",
      "<tensorflow.python.keras.layers.pooling.MaxPooling2D object at 0x0000024E14183D60>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E17E335B0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E17E339D0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E17EDE790>\n",
      "<tensorflow.python.keras.layers.pooling.MaxPooling2D object at 0x0000024E17EB40D0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E0DEE1E50>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E18072AF0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E18097820>\n",
      "<tensorflow.python.keras.layers.pooling.MaxPooling2D object at 0x0000024E17EB4B80>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E180ED1C0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E180B59A0>\n",
      "<tensorflow.python.keras.layers.convolutional.Conv2D object at 0x0000024E180E4AC0>\n",
      "<tensorflow.python.keras.layers.pooling.MaxPooling2D object at 0x0000024E1421AE80>\n"
     ]
    }
   ],
   "source": [
    "print ('Performing Transfer Learning')\n",
    "\n",
    "for layer in base_model.layers: \n",
    "    print(layer)\n",
    "    layer.trainable = False\n",
    "    \n",
    "trans_model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"functional_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         [(None, 64, 64, 3)]       0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 64, 64, 64)        1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 64, 64, 64)        36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 32, 32, 64)        0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 32, 32, 128)       73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 32, 32, 128)       147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 16, 16, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 16, 16, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 16, 16, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 16, 16, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 8, 8, 256)         0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 8, 8, 512)         1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 8, 8, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 4, 4, 512)         0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 4, 4, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 2, 2, 512)         0         \n",
      "_________________________________________________________________\n",
      "flatten_4 (Flatten)          (None, 2048)              0         \n",
      "_________________________________________________________________\n",
      "dense_12 (Dense)             (None, 64)                131136    \n",
      "_________________________________________________________________\n",
      "dense_13 (Dense)             (None, 64)                4160      \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 14,850,049\n",
      "Trainable params: 135,361\n",
      "Non-trainable params: 14,714,688\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "trans_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "20/20 [==============================] - 7s 338ms/step - loss: 0.6091 - accuracy: 0.6675 - val_loss: 0.4120 - val_accuracy: 0.8600\n",
      "Epoch 2/100\n",
      "20/20 [==============================] - 7s 335ms/step - loss: 0.4413 - accuracy: 0.7900 - val_loss: 0.2825 - val_accuracy: 0.8900\n",
      "Epoch 3/100\n",
      "20/20 [==============================] - 7s 368ms/step - loss: 0.3161 - accuracy: 0.8725 - val_loss: 0.2466 - val_accuracy: 0.9100\n",
      "Epoch 4/100\n",
      "20/20 [==============================] - 8s 395ms/step - loss: 0.2323 - accuracy: 0.9025 - val_loss: 0.1767 - val_accuracy: 0.9200\n",
      "Epoch 5/100\n",
      "20/20 [==============================] - 8s 396ms/step - loss: 0.1874 - accuracy: 0.9300 - val_loss: 0.1535 - val_accuracy: 0.9500\n",
      "Epoch 6/100\n",
      "20/20 [==============================] - 8s 400ms/step - loss: 0.1742 - accuracy: 0.9425 - val_loss: 0.1404 - val_accuracy: 0.9400\n",
      "Epoch 7/100\n",
      "20/20 [==============================] - 8s 417ms/step - loss: 0.1271 - accuracy: 0.9575 - val_loss: 0.1946 - val_accuracy: 0.9400\n",
      "Epoch 8/100\n",
      "20/20 [==============================] - 9s 450ms/step - loss: 0.1202 - accuracy: 0.9625 - val_loss: 0.1308 - val_accuracy: 0.9600\n",
      "Epoch 9/100\n",
      "20/20 [==============================] - 8s 420ms/step - loss: 0.1021 - accuracy: 0.9675 - val_loss: 0.1617 - val_accuracy: 0.9600\n",
      "Epoch 10/100\n",
      "20/20 [==============================] - 10s 498ms/step - loss: 0.0805 - accuracy: 0.9825 - val_loss: 0.1382 - val_accuracy: 0.9400\n",
      "Epoch 11/100\n",
      "20/20 [==============================] - 10s 483ms/step - loss: 0.0493 - accuracy: 0.9950 - val_loss: 0.1203 - val_accuracy: 0.9200\n",
      "Epoch 12/100\n",
      "20/20 [==============================] - 8s 422ms/step - loss: 0.0498 - accuracy: 0.9850 - val_loss: 0.1183 - val_accuracy: 0.9300\n",
      "Epoch 13/100\n",
      "20/20 [==============================] - 9s 454ms/step - loss: 0.0540 - accuracy: 0.9875 - val_loss: 0.1974 - val_accuracy: 0.9300\n",
      "Epoch 14/100\n",
      "20/20 [==============================] - 8s 418ms/step - loss: 0.0993 - accuracy: 0.9700 - val_loss: 0.1545 - val_accuracy: 0.9500\n",
      "Epoch 15/100\n",
      "20/20 [==============================] - 8s 409ms/step - loss: 0.0322 - accuracy: 0.9950 - val_loss: 0.1226 - val_accuracy: 0.9300\n"
     ]
    }
   ],
   "source": [
    "early_stop = callbacks.EarlyStopping(monitor= 'val_loss', patience = 3)\n",
    "\n",
    "hist_trans1 = trans_model.fit(X_train, y_train, epochs=100,batch_size=batch_size, validation_data = (X_test, y_test), callbacks=[early_stop])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "epoch_list = list (range (1, len(hist_trans1.history['accuracy'])+1))\n",
    "plt.figure(figsize =(10,10))\n",
    "plt.plot(epoch_list, hist_trans1.history['accuracy'])\n",
    "plt.plot(epoch_list, hist_trans1.history['val_accuracy'])\n",
    "plt.legend(['Training accuracy', 'Accuracy'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Both Image Data Generator and Transfer learning help develop a more accurate model for prediction. "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}