cse474_assignment2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import csv\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import tensorflow.keras.callbacks as callbacks\n",
    "from sklearn import preprocessing as pp\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.utils import to_categorical \n",
    "from tensorflow.keras.optimizers import SGD, Adam, Nadam, Adamax, Adadelta, Adagrad, Ftrl, RMSprop, Optimizer\n",
    "from keras_visualizer import visualizer\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = \"/Users/romi_sai/Desktop/College/Senior/Summer2021/CSE474/Assignments/Assignment2/diabetes.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#opening csv & reading in rows \n",
    "#some parts of this code was found on Python Docs (https://docs.python.org/3/library/csv.html)\n",
    "with open(filename, newline='') as csvfile:\n",
    "    lines = csv.reader(csvfile)\n",
    "    rows=[]\n",
    "    for row in lines:\n",
    "        rows.append(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Converting the age, bmi, glucose and outcome to lists for easy access\n",
    "pregnancies=[]\n",
    "glucose=[]\n",
    "bp=[]\n",
    "skinthickness=[]\n",
    "insulin=[]\n",
    "bmi=[]\n",
    "pedigree=[]\n",
    "age=[]\n",
    "out=[]\n",
    "for thing in rows:\n",
    "    pregnancies.append(thing[0])\n",
    "    glucose.append(thing[1])\n",
    "    bp.append(thing[2])\n",
    "    skinthickness.append(thing[3])\n",
    "    insulin.append(thing[4])\n",
    "    bmi.append(thing[5])\n",
    "    pedigree.append(thing[6])\n",
    "    age.append(thing[7])\n",
    "    out.append(thing[8])\n",
    "    \n",
    "    \n",
    "pregnancies.remove('Pregnancies')\n",
    "glucose.remove('Glucose')\n",
    "bp.remove(\"BloodPressure\")\n",
    "skinthickness.remove(\"SkinThickness\")\n",
    "insulin.remove(\"Insulin\")\n",
    "bmi.remove(\"BMI\")\n",
    "pedigree.remove(\"DiabetesPedigreeFunction\")\n",
    "age.remove(\"Age\")\n",
    "out.remove(\"Outcome\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prep: Extracting, Splitting & Normalizing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#defining the train and test datasets for the features: age, bmi and glucose for X and charges for Y\n",
    "#train_bias=[]\n",
    "train_gluc=[]\n",
    "train_bmi=[]\n",
    "train_age=[]\n",
    "train_out=[]\n",
    "#test_bias=[]\n",
    "test_age=[]\n",
    "test_bmi=[]\n",
    "test_gluc=[]\n",
    "test_out=[]\n",
    "#80%\n",
    "for i in range(0, 614, 1):\n",
    "   # train_bias.append(1)\n",
    "    train_gluc.append(glucose[i])\n",
    "    train_bmi.append(bmi[i])\n",
    "    train_age.append(age[i])\n",
    "    train_out.append(out[i])\n",
    "\n",
    "#20%   \n",
    "for j in range(614, 768, 1):\n",
    "    #print(j)\n",
    "    #test_bias.append(1)\n",
    "    test_age.append(age[j])\n",
    "    test_bmi.append(bmi[j])\n",
    "    test_gluc.append(glucose[j])\n",
    "    test_out.append(out[j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "###Converting the lists from strings to integers and floats\n",
    "##This function was found on the numpy docs (https://numpy.org/doc/stable/reference/generated/numpy.matrix.html)\n",
    "Xtrain = np.matrix((train_gluc, train_bmi,train_age), dtype=float)\n",
    "Xtrain = Xtrain.transpose()\n",
    "\n",
    "Xtest = np.matrix((test_gluc, test_bmi, test_age), dtype=float)\n",
    "Xtest = Xtest.transpose()\n",
    "\n",
    "Ytrain = np.matrix((train_out), dtype=float)\n",
    "Ytrain = Ytrain.transpose()\n",
    "\n",
    "Ytest = np.matrix((test_out), dtype=float)\n",
    "Ytest = Ytest.transpose()\n",
    "\n",
    "#print(Xtest.shape)\n",
    "#Xtest.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Normalizing the data\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(Xtrain)\n",
    "Xtrain = scaler.transform(Xtrain)\n",
    "\n",
    "scaler.fit(Xtest)\n",
    "Xtest = scaler.transform(Xtest)\n",
    "\n",
    "## not normalizing the Ytest and Ytrain bc only the features need to be normalized for us to predict the right outcomes. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1: Implementing a Neural Network with One Hidden Layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_59\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_118 (Dense)            (None, 90)                360       \n",
      "_________________________________________________________________\n",
      "dense_119 (Dense)            (None, 2)                 182       \n",
      "=================================================================\n",
      "Total params: 542\n",
      "Trainable params: 542\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Epoch 1/80\n",
      "11/11 [==============================] - 0s 11ms/step - loss: 0.6879 - accuracy: 0.5000 - val_loss: 0.6646 - val_accuracy: 0.5779\n",
      "Epoch 2/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6821 - accuracy: 0.5309 - val_loss: 0.6592 - val_accuracy: 0.6234\n",
      "Epoch 3/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6766 - accuracy: 0.5700 - val_loss: 0.6543 - val_accuracy: 0.6818\n",
      "Epoch 4/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6717 - accuracy: 0.6026 - val_loss: 0.6499 - val_accuracy: 0.7208\n",
      "Epoch 5/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6673 - accuracy: 0.6319 - val_loss: 0.6459 - val_accuracy: 0.7273\n",
      "Epoch 6/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6630 - accuracy: 0.6482 - val_loss: 0.6421 - val_accuracy: 0.7273\n",
      "Epoch 7/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6590 - accuracy: 0.6661 - val_loss: 0.6384 - val_accuracy: 0.7273\n",
      "Epoch 8/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6553 - accuracy: 0.6792 - val_loss: 0.6348 - val_accuracy: 0.7273\n",
      "Epoch 9/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6516 - accuracy: 0.6857 - val_loss: 0.6313 - val_accuracy: 0.7338\n",
      "Epoch 10/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6480 - accuracy: 0.6889 - val_loss: 0.6278 - val_accuracy: 0.7403\n",
      "Epoch 11/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6444 - accuracy: 0.6906 - val_loss: 0.6245 - val_accuracy: 0.7468\n",
      "Epoch 12/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6409 - accuracy: 0.6971 - val_loss: 0.6213 - val_accuracy: 0.7468\n",
      "Epoch 13/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6376 - accuracy: 0.6971 - val_loss: 0.6180 - val_accuracy: 0.7403\n",
      "Epoch 14/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6343 - accuracy: 0.7068 - val_loss: 0.6148 - val_accuracy: 0.7403\n",
      "Epoch 15/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6311 - accuracy: 0.7150 - val_loss: 0.6117 - val_accuracy: 0.7597\n",
      "Epoch 16/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6278 - accuracy: 0.7215 - val_loss: 0.6088 - val_accuracy: 0.7597\n",
      "Epoch 17/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6248 - accuracy: 0.7280 - val_loss: 0.6059 - val_accuracy: 0.7532\n",
      "Epoch 18/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.6219 - accuracy: 0.7248 - val_loss: 0.6030 - val_accuracy: 0.7532\n",
      "Epoch 19/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6189 - accuracy: 0.7264 - val_loss: 0.6003 - val_accuracy: 0.7532\n",
      "Epoch 20/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6161 - accuracy: 0.7248 - val_loss: 0.5975 - val_accuracy: 0.7662\n",
      "Epoch 21/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6132 - accuracy: 0.7296 - val_loss: 0.5947 - val_accuracy: 0.7727\n",
      "Epoch 22/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6103 - accuracy: 0.7296 - val_loss: 0.5921 - val_accuracy: 0.7792\n",
      "Epoch 23/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6076 - accuracy: 0.7296 - val_loss: 0.5895 - val_accuracy: 0.7792\n",
      "Epoch 24/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6049 - accuracy: 0.7296 - val_loss: 0.5870 - val_accuracy: 0.7857\n",
      "Epoch 25/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.6022 - accuracy: 0.7329 - val_loss: 0.5845 - val_accuracy: 0.7792\n",
      "Epoch 26/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5997 - accuracy: 0.7313 - val_loss: 0.5822 - val_accuracy: 0.7792\n",
      "Epoch 27/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5971 - accuracy: 0.7313 - val_loss: 0.5797 - val_accuracy: 0.7857\n",
      "Epoch 28/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5946 - accuracy: 0.7329 - val_loss: 0.5775 - val_accuracy: 0.7922\n",
      "Epoch 29/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.5923 - accuracy: 0.7329 - val_loss: 0.5752 - val_accuracy: 0.7922\n",
      "Epoch 30/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5899 - accuracy: 0.7345 - val_loss: 0.5731 - val_accuracy: 0.7922\n",
      "Epoch 31/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5876 - accuracy: 0.7362 - val_loss: 0.5710 - val_accuracy: 0.7922\n",
      "Epoch 32/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5854 - accuracy: 0.7378 - val_loss: 0.5688 - val_accuracy: 0.7922\n",
      "Epoch 33/80\n",
      "11/11 [==============================] - 0s 3ms/step - loss: 0.5830 - accuracy: 0.7362 - val_loss: 0.5668 - val_accuracy: 0.7922\n",
      "Epoch 34/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5808 - accuracy: 0.7378 - val_loss: 0.5647 - val_accuracy: 0.7987\n",
      "Epoch 35/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5787 - accuracy: 0.7362 - val_loss: 0.5628 - val_accuracy: 0.7987\n",
      "Epoch 36/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5766 - accuracy: 0.7394 - val_loss: 0.5608 - val_accuracy: 0.8117\n",
      "Epoch 37/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5745 - accuracy: 0.7394 - val_loss: 0.5590 - val_accuracy: 0.8117\n",
      "Epoch 38/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5726 - accuracy: 0.7394 - val_loss: 0.5572 - val_accuracy: 0.8117\n",
      "Epoch 39/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5706 - accuracy: 0.7427 - val_loss: 0.5554 - val_accuracy: 0.8117\n",
      "Epoch 40/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5687 - accuracy: 0.7427 - val_loss: 0.5537 - val_accuracy: 0.8117\n",
      "Epoch 41/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5670 - accuracy: 0.7459 - val_loss: 0.5520 - val_accuracy: 0.7987\n",
      "Epoch 42/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5651 - accuracy: 0.7476 - val_loss: 0.5504 - val_accuracy: 0.7987\n",
      "Epoch 43/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5633 - accuracy: 0.7492 - val_loss: 0.5488 - val_accuracy: 0.7987\n",
      "Epoch 44/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5614 - accuracy: 0.7492 - val_loss: 0.5473 - val_accuracy: 0.7987\n",
      "Epoch 45/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5597 - accuracy: 0.7476 - val_loss: 0.5458 - val_accuracy: 0.7987\n",
      "Epoch 46/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5580 - accuracy: 0.7459 - val_loss: 0.5443 - val_accuracy: 0.7987\n",
      "Epoch 47/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5563 - accuracy: 0.7443 - val_loss: 0.5427 - val_accuracy: 0.7987\n",
      "Epoch 48/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5546 - accuracy: 0.7443 - val_loss: 0.5412 - val_accuracy: 0.7987\n",
      "Epoch 49/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5529 - accuracy: 0.7443 - val_loss: 0.5398 - val_accuracy: 0.7987\n",
      "Epoch 50/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5513 - accuracy: 0.7427 - val_loss: 0.5385 - val_accuracy: 0.7922\n",
      "Epoch 51/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5498 - accuracy: 0.7459 - val_loss: 0.5371 - val_accuracy: 0.7922\n",
      "Epoch 52/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5483 - accuracy: 0.7459 - val_loss: 0.5357 - val_accuracy: 0.7922\n",
      "Epoch 53/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5467 - accuracy: 0.7459 - val_loss: 0.5345 - val_accuracy: 0.7922\n",
      "Epoch 54/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5454 - accuracy: 0.7459 - val_loss: 0.5332 - val_accuracy: 0.7922\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 55/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5439 - accuracy: 0.7459 - val_loss: 0.5319 - val_accuracy: 0.7922\n",
      "Epoch 56/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5424 - accuracy: 0.7492 - val_loss: 0.5307 - val_accuracy: 0.7922\n",
      "Epoch 57/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5411 - accuracy: 0.7508 - val_loss: 0.5295 - val_accuracy: 0.7922\n",
      "Epoch 58/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5398 - accuracy: 0.7492 - val_loss: 0.5283 - val_accuracy: 0.7987\n",
      "Epoch 59/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5385 - accuracy: 0.7524 - val_loss: 0.5272 - val_accuracy: 0.7987\n",
      "Epoch 60/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5372 - accuracy: 0.7541 - val_loss: 0.5261 - val_accuracy: 0.7987\n",
      "Epoch 61/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5359 - accuracy: 0.7557 - val_loss: 0.5249 - val_accuracy: 0.7987\n",
      "Epoch 62/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5347 - accuracy: 0.7573 - val_loss: 0.5239 - val_accuracy: 0.7987\n",
      "Epoch 63/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5334 - accuracy: 0.7557 - val_loss: 0.5229 - val_accuracy: 0.7987\n",
      "Epoch 64/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5322 - accuracy: 0.7557 - val_loss: 0.5219 - val_accuracy: 0.7987\n",
      "Epoch 65/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5310 - accuracy: 0.7557 - val_loss: 0.5209 - val_accuracy: 0.7987\n",
      "Epoch 66/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5299 - accuracy: 0.7541 - val_loss: 0.5198 - val_accuracy: 0.7987\n",
      "Epoch 67/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5288 - accuracy: 0.7541 - val_loss: 0.5189 - val_accuracy: 0.7987\n",
      "Epoch 68/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5276 - accuracy: 0.7557 - val_loss: 0.5179 - val_accuracy: 0.7987\n",
      "Epoch 69/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5265 - accuracy: 0.7557 - val_loss: 0.5170 - val_accuracy: 0.7987\n",
      "Epoch 70/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5254 - accuracy: 0.7557 - val_loss: 0.5161 - val_accuracy: 0.7987\n",
      "Epoch 71/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5243 - accuracy: 0.7557 - val_loss: 0.5152 - val_accuracy: 0.7987\n",
      "Epoch 72/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5233 - accuracy: 0.7557 - val_loss: 0.5143 - val_accuracy: 0.7987\n",
      "Epoch 73/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5223 - accuracy: 0.7541 - val_loss: 0.5136 - val_accuracy: 0.7987\n",
      "Epoch 74/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5214 - accuracy: 0.7541 - val_loss: 0.5128 - val_accuracy: 0.7987\n",
      "Epoch 75/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5205 - accuracy: 0.7541 - val_loss: 0.5121 - val_accuracy: 0.7987\n",
      "Epoch 76/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5195 - accuracy: 0.7541 - val_loss: 0.5113 - val_accuracy: 0.7987\n",
      "Epoch 77/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5186 - accuracy: 0.7541 - val_loss: 0.5105 - val_accuracy: 0.7987\n",
      "Epoch 78/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5177 - accuracy: 0.7541 - val_loss: 0.5099 - val_accuracy: 0.7987\n",
      "Epoch 79/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5169 - accuracy: 0.7541 - val_loss: 0.5091 - val_accuracy: 0.7987\n",
      "Epoch 80/80\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 0.5160 - accuracy: 0.7541 - val_loss: 0.5084 - val_accuracy: 0.7987\n",
      "\n",
      "\n",
      "\n",
      " TESTING LOSS vs ACCURACY\n",
      "5/5 [==============================] - 0s 757us/step - loss: 0.5084 - accuracy: 0.7987\n",
      "\n",
      " Loss 0.5084162354469299 Accuracy: 0.798701286315918\n",
      "\n",
      "\n",
      "\n",
      " TRAINING LOSS vs ACCURACY\n",
      "20/20 [==============================] - 0s 565us/step - loss: 0.5154 - accuracy: 0.7541\n",
      "\n",
      " Loss 0.515447735786438 Accuracy: 0.7540716528892517\n"
     ]
    }
   ],
   "source": [
    "# Creating the model.\n",
    "model = Sequential()\n",
    "\n",
    "# Adding layers to the model using Dense(#neurons, activation func, dimensions of the input layer)\n",
    "model.add(Dense(90, activation='relu', input_dim=3))  # Hidden layer.\n",
    "\n",
    "#Using sigmoid because it's a binary classification\n",
    "model.add(Dense(2, activation='sigmoid'))  # Output layer.\n",
    "#visualizer(model, format='png', view=True)\n",
    "\n",
    "model.compile(optimizer=Adamax(learning_rate=0.0001), loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "print(model.summary())\n",
    "\n",
    "# Training the model.\n",
    "#lowering the batch_size is increasing accuracy!\n",
    "history = model.fit(Xtrain, to_categorical(Ytrain),\n",
    "                    validation_data=(Xtest, to_categorical(Ytest)), epochs=80, batch_size=60)\n",
    "\n",
    "# Evaluating the model - TEST DATA\n",
    "print('\\n\\n\\n TESTING LOSS vs ACCURACY')\n",
    "loss, accuracy = model.evaluate(Xtest, to_categorical(Ytest))\n",
    "print('\\n Loss', loss, 'Accuracy:', accuracy)\n",
    "\n",
    "\n",
    "# Evaluating model using TRAINING DATA\n",
    "print('\\n\\n\\n TRAINING LOSS vs ACCURACY')\n",
    "loss_tr, acc = model.evaluate(Xtrain, to_categorical(Ytrain))\n",
    "print('\\n Loss', loss_tr, 'Accuracy:', acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting the graphs "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train vs Validation Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the accuracy and loss.\n",
    "\n",
    "# Accuracy\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('1 Hidden Layer: Training and Testing Error')\n",
    "plt.ylabel('Error')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Training Accuracy', 'Testing Accuracy'], loc='upper right')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train vs Validation Loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Loss\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('1 Hidden Layer: Training and Testing Loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Training Loss', 'Testing Loss'], loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 2: Neural Network with Three Hidden Layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_22\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_88 (Dense)             (None, 512)               2048      \n",
      "_________________________________________________________________\n",
      "dense_89 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dense_90 (Dense)             (None, 128)               32896     \n",
      "_________________________________________________________________\n",
      "dense_91 (Dense)             (None, 2)                 258       \n",
      "=================================================================\n",
      "Total params: 166,530\n",
      "Trainable params: 166,530\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Epoch 1/25\n",
      "10/10 [==============================] - 0s 15ms/step - loss: 17.6685 - accuracy: 0.5619 - val_loss: 18.9813 - val_accuracy: 0.6883\n",
      "Epoch 2/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 13.8145 - accuracy: 0.6498 - val_loss: 7.8532 - val_accuracy: 0.6623\n",
      "Epoch 3/25\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 5.9656 - accuracy: 0.7134 - val_loss: 4.3044 - val_accuracy: 0.7922\n",
      "Epoch 4/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 3.1397 - accuracy: 0.7280 - val_loss: 2.0144 - val_accuracy: 0.7597\n",
      "Epoch 5/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 1.5364 - accuracy: 0.7671 - val_loss: 1.1189 - val_accuracy: 0.7597\n",
      "Epoch 6/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.9343 - accuracy: 0.7541 - val_loss: 0.7694 - val_accuracy: 0.7468\n",
      "Epoch 7/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.6851 - accuracy: 0.7638 - val_loss: 0.6066 - val_accuracy: 0.7922\n",
      "Epoch 8/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5874 - accuracy: 0.7492 - val_loss: 0.6302 - val_accuracy: 0.7403\n",
      "Epoch 9/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5902 - accuracy: 0.7476 - val_loss: 0.5754 - val_accuracy: 0.7597\n",
      "Epoch 10/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5484 - accuracy: 0.7557 - val_loss: 0.5448 - val_accuracy: 0.7792\n",
      "Epoch 11/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5690 - accuracy: 0.7443 - val_loss: 0.6036 - val_accuracy: 0.7208\n",
      "Epoch 12/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5603 - accuracy: 0.7378 - val_loss: 0.5339 - val_accuracy: 0.7857\n",
      "Epoch 13/25\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.5233 - accuracy: 0.7671 - val_loss: 0.5410 - val_accuracy: 0.7597\n",
      "Epoch 14/25\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.5312 - accuracy: 0.7459 - val_loss: 0.5277 - val_accuracy: 0.7857\n",
      "Epoch 15/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5265 - accuracy: 0.7638 - val_loss: 0.5451 - val_accuracy: 0.7468\n",
      "Epoch 16/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5215 - accuracy: 0.7476 - val_loss: 0.5556 - val_accuracy: 0.7532\n",
      "Epoch 17/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5371 - accuracy: 0.7427 - val_loss: 0.5494 - val_accuracy: 0.7597\n",
      "Epoch 18/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5648 - accuracy: 0.7459 - val_loss: 0.5586 - val_accuracy: 0.7662\n",
      "Epoch 19/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5631 - accuracy: 0.7362 - val_loss: 0.5600 - val_accuracy: 0.7662\n",
      "Epoch 20/25\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.5423 - accuracy: 0.7378 - val_loss: 0.5279 - val_accuracy: 0.7922\n",
      "Epoch 21/25\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.5308 - accuracy: 0.7687 - val_loss: 0.5230 - val_accuracy: 0.7857\n",
      "Epoch 22/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5160 - accuracy: 0.7704 - val_loss: 0.5286 - val_accuracy: 0.7727\n",
      "Epoch 23/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5153 - accuracy: 0.7769 - val_loss: 0.5268 - val_accuracy: 0.7727\n",
      "Epoch 24/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5153 - accuracy: 0.7704 - val_loss: 0.5224 - val_accuracy: 0.7727\n",
      "Epoch 25/25\n",
      "10/10 [==============================] - 0s 4ms/step - loss: 0.5133 - accuracy: 0.7736 - val_loss: 0.5214 - val_accuracy: 0.7792\n",
      "\n",
      "\n",
      "\n",
      " TESTING LOSS vs ACCURACY\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 0.5214 - accuracy: 0.7792\n",
      "\n",
      " Loss 0.5214400291442871 Accuracy: 0.7792207598686218\n",
      "\n",
      "\n",
      "\n",
      " TRAINING LOSS vs ACCURACY\n",
      "20/20 [==============================] - 0s 795us/step - loss: 0.5056 - accuracy: 0.7785\n",
      "\n",
      " Loss 0.505594789981842 Accuracy: 0.7785016298294067\n"
     ]
    }
   ],
   "source": [
    "# Neural Network with Three Hidden Layers.\n",
    "\n",
    "# Creating the model.\n",
    "model = Sequential()\n",
    "model.add(Dense(512, activation='relu', input_dim=3, kernel_regularizer='l2'))\n",
    "model.add(Dense(256, activation='relu', kernel_regularizer='l2'))\n",
    "model.add(Dense(128, activation='relu', kernel_regularizer='l2'))\n",
    "model.add(Dense(2, activation='sigmoid'))\n",
    "#visualizer(model, format='png', view=True)\n",
    "\n",
    "#using keras, rlr on plateau\n",
    "model.compile(optimizer=Adam(learning_rate=0.1), loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "print(model.summary())\n",
    "\n",
    "# Training the model.\n",
    "reduce_lr = callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.2, patience=5, min_lr=0.00001)\n",
    "#history = model.fit(X_train, to_categorical(Ytrain), callbacks=[reduce_lr])\n",
    "\n",
    "history = model.fit(Xtrain, to_categorical(Ytrain),\n",
    "                    validation_data=(Xtest, to_categorical(Ytest)), epochs=25, batch_size=64, callbacks=[reduce_lr])\n",
    "\n",
    "# Evaluating the model - TEST DATA\n",
    "print('\\n\\n\\n TESTING LOSS vs ACCURACY')\n",
    "l, a = model.evaluate(Xtest, to_categorical(Ytest))\n",
    "print('\\n Loss', l, 'Accuracy:', a)\n",
    "\n",
    "\n",
    "# Evaluating model using TRAINING DATA\n",
    "print('\\n\\n\\n TRAINING LOSS vs ACCURACY')\n",
    "ll, ac = model.evaluate(Xtrain, to_categorical(Ytrain))\n",
    "print('\\n Loss', ll, 'Accuracy:', ac)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting the graphs "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train vs Validation Accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the accuracy and loss.\n",
    "\n",
    "# Accuracy\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('3 Hidden Layers: Training and Testing Error')\n",
    "plt.ylabel('Error')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Training Accuracy', 'Testing Accuracy'], loc='upper right')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train vs Validation Loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# Loss\n",
    "plt.figure(figsize=(15, 10))\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('3 Hidden Layers: Training and Testing Loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Training Loss', 'Testing Loss'], loc='upper right')\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}