Mushroom_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 380,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.ensemble import RandomForestClassifier as RFC\n",
    "from sklearn.tree import DecisionTreeClassifier as DTC\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "import datetime\n",
    "import pandas as pd\n",
    "from sklearn import preprocessing as pp\n",
    "from sklearn.linear_model import LogisticRegression as LR\n",
    "from sklearn.neighbors import KNeighborsClassifier as KNN\n",
    "\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation\n",
    "from keras.utils import np_utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8124, 23)\n"
     ]
    }
   ],
   "source": [
    "ds = pd.read_csv('mushrooms.csv')\n",
    "dat = ds.values\n",
    "\n",
    "print dat.shape\n",
    "\n",
    "headers = list(ds.columns.values) #store features of mushrooms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#IGNORE THIS CELL\n",
    "\n",
    "arr = np.array([1,2,3])\n",
    "b = np.array([4, 5, 6])\n",
    "\n",
    "arr = np.vstack((arr, b))\n",
    "#q = np.concatenate((arr,b), axis=0)\n",
    "#print arr\n",
    "\n",
    "le = pp.LabelEncoder()\n",
    "le.fit(dat[:, 0])\n",
    "y = le.transform(dat[:, 0])\n",
    "le1 = pp.LabelEncoder()\n",
    "le1.fit(dat[:, 1])\n",
    "z = le1.transform(dat[:, 1])\n",
    "le2 = pp.LabelEncoder()\n",
    "le2.fit(dat[:, 2])\n",
    "zz = le2.transform(dat[:, 2])\n",
    "\n",
    "# y = np.vstack(y, z)\n",
    "# z = np.vstack(z)\n",
    "# y = np.vstack(y)\n",
    "qqw = np.vstack((y, z))\n",
    "qqw = np.vstack((qqw, zz))\n",
    "#print qqw.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 383,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8124, 23)\n"
     ]
    }
   ],
   "source": [
    "#Data Preprocessing\n",
    "l = pp.LabelEncoder()\n",
    "l.fit(dat[:, 0])\n",
    "dataa = l.transform(dat[:, 0])\n",
    "\n",
    "for ix in range(1, dat.shape[1]):\n",
    "    le = pp.LabelEncoder()\n",
    "    le.fit(dat[:, ix])\n",
    "    y = le.transform(dat[:, ix])\n",
    "    dataa = np.vstack((dataa , y))\n",
    "    \n",
    "data = dataa.T\n",
    "\n",
    "cate = data[:, 0] #One hot encoding for Neural Network implementation\n",
    "\n",
    "print data.shape ##FINALLLYY O YEAH!!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 384,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(6499, 22) (6499,)\n",
      "(1625, 22) (1625,)\n"
     ]
    }
   ],
   "source": [
    "split = int(0.80 * data.shape[0])\n",
    "\n",
    "x_train = data[:split , 1:]\n",
    "y_train = data[:split, 0]\n",
    "\n",
    "x_test = data[split: , 1:]\n",
    "y_test = data[split: , 0]\n",
    "\n",
    "print x_train.shape, y_train.shape\n",
    "print x_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> Decision Tree Implementation </h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "acc = []\n",
    "ans = []\n",
    "\n",
    "for ix in range(10):\n",
    "    dt = DTC()\n",
    "    \n",
    "    start = datetime.datetime.now()\n",
    "    dt.fit(x_train, y_train)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Training Time : \", end-start\n",
    "\n",
    "    start = datetime.datetime.now()\n",
    "    score = dt.score(x_test, y_test)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Testing Time : \", end-start\n",
    "\n",
    "    #print \"Accurcy : \", score*100 \n",
    "    acc.append(score*100)\n",
    "    tem = dt.feature_importances_\n",
    "    ans.append(tem)\n",
    "    #print \"\\n\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 386,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features most indicative of a poisonous mushroom wrt Decision Tree Model :  spore-print-color\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjEAAAF9CAYAAAD1HTl+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n15k4Mt3LFgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x111217fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp = [] \n",
    "for ix in range(0, len(ans)):\n",
    "    temp.append(np.argmax(ans[ix]))\n",
    "    \n",
    "mode = max(set(temp), key=temp.count) #find mode for features importance in decision trees\n",
    "print \"Features most indicative of a poisonous mushroom wrt Decision Tree Model : \", headers[mode+1]\n",
    "\n",
    "plt.figure(0)\n",
    "plt.suptitle('Decision Tree Plot', fontsize=10)\n",
    "plt.plot(acc, '-o')\n",
    "plt.ylabel('Accuracy', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> Random Forest Implementation </h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 387,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "est = [] #taking variable estimators\n",
    "acc = []\n",
    "ans = []\n",
    "for iy in range(2, 200, 20):\n",
    "    est.append(iy)\n",
    "\n",
    "for ix in range(len(est)):\n",
    "    rf = RFC(n_estimators=est[ix], n_jobs=2)\n",
    "    #print \"No. of Decision Trees : \", est[ix]\n",
    "    \n",
    "    start = datetime.datetime.now()\n",
    "    rf.fit(x_train, y_train)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Training Time : \", end-start\n",
    "\n",
    "    start = datetime.datetime.now()\n",
    "    score = rf.score(x_test, y_test)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Testing Time : \", end-start\n",
    "\n",
    "    #print \"Accurcy : \", score*100 \n",
    "    acc.append(score*100)\n",
    "    tem = rf.feature_importances_\n",
    "    ans.append(tem)\n",
    "    #print \"\\n\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features most indicative of a poisonous mushroom wrt Random Forest Model :  odor\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjoAAAGYCAYAAABLdEi4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWd1yomNmZmZ16/8BB5rIgL3b8zkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x111b2cc90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp = []\n",
    "\n",
    "for ix in range(0, len(ans)):\n",
    "    temp.append(np.argmax(ans[ix]))\n",
    "\n",
    "mode = max(set(temp), key=temp.count) #find mode for features importance in variable estimators\n",
    "print \"Features most indicative of a poisonous mushroom wrt Random Forest Model : \", headers[mode+1]\n",
    "\n",
    "plt.figure(1)\n",
    "plt.suptitle('Random Forest Plot', fontsize=10)\n",
    "plt.plot(est, acc, '-o')\n",
    "plt.xlabel('Estimators', fontsize=16)\n",
    "plt.ylabel('Accuracy', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> Logistic Regression Implementation </h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 389,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Time :  0:00:00.108130\n",
      "Testing Time :  0:00:00.000907\n",
      "Accurcy :  89.4153846154\n"
     ]
    }
   ],
   "source": [
    "lr = LR(n_jobs=-1)\n",
    "\n",
    "start = datetime.datetime.now()\n",
    "lr.fit(x_train, y_train)\n",
    "end = datetime.datetime.now()\n",
    "\n",
    "print \"Training Time : \", end-start\n",
    "\n",
    "start = datetime.datetime.now()\n",
    "score = lr.score(x_test, y_test)\n",
    "end = datetime.datetime.now()\n",
    "\n",
    "print \"Testing Time : \", end-start\n",
    "\n",
    "print \"Accurcy : \", score*100 \n",
    "\n",
    "features = dt.feature_importances_\n",
    "#print \"\\n\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features most indicative of a poisonous mushroom wrt Logistic Regression Model :  spore-print-color\n"
     ]
    }
   ],
   "source": [
    "temp = np.argmax(features)\n",
    "\n",
    "print \"Features most indicative of a poisonous mushroom wrt Logistic Regression Model : \", headers[temp+1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> kNN Implementation </h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 391,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "acc = []\n",
    "ans = []\n",
    "neighbours = []\n",
    "\n",
    "for ix in range(3, 200, 20):\n",
    "    neighbours.append(ix)\n",
    "\n",
    "for ix in range(len(neighbours)):\n",
    "    knn = KNN(n_neighbors=neighbours[ix], n_jobs=-1)\n",
    "    \n",
    "    start = datetime.datetime.now()\n",
    "    knn.fit(x_train, y_train)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Training Time : \", end-start\n",
    "\n",
    "    start = datetime.datetime.now()\n",
    "    score = knn.score(x_test, y_test)\n",
    "    end = datetime.datetime.now()\n",
    "\n",
    "    #print \"Testing Time : \", end-start\n",
    "\n",
    "    #print \"Accurcy : \", score*100 \n",
    "    acc.append(score*100)\n",
    "    temp = dt.feature_importances_\n",
    "    ans.append(temp)\n",
    "    #print \"\\n\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 392,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features most indicative of a poisonous mushroom wrt kNN :  spore-print-color\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiQAAAGYCAYAAAByE/njAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZqX7H2qqzRTnTH1tAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1112b14d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp = []\n",
    "\n",
    "for ix in range(0, len(ans)):\n",
    "    temp.append(np.argmax(ans[ix]))\n",
    "\n",
    "mode = max(set(temp), key=temp.count) #find mode for features importance in variable estimators\n",
    "print \"Features most indicative of a poisonous mushroom wrt kNN : \", headers[mode+1]\n",
    "\n",
    "plt.figure(2)\n",
    "plt.suptitle('k-Nearest Neighbour Plot', fontsize=10)\n",
    "plt.plot(neighbours, acc, '-o')\n",
    "plt.xlabel('Neighbours', fontsize=16)\n",
    "plt.ylabel('Accuracy', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3> Neural Network Implementation Using Keras </h3> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 393,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(6499, 22) (1625, 22)\n",
      "(6499,) (1625,)\n",
      "(8124, 23)\n"
     ]
    }
   ],
   "source": [
    "y = np_utils.to_categorical(cate)\n",
    "Y_train = y[:split]\n",
    "Y_test = y[split:]\n",
    "\n",
    "print x_train.shape, x_test.shape\n",
    "print y_train.shape, y_test.shape\n",
    "print data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 394,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "____________________________________________________________________________________________________\n",
      "Layer (type)                     Output Shape          Param #     Connected to                     \n",
      "====================================================================================================\n",
      "dense_46 (Dense)                 (None, 11)            253         dense_input_16[0][0]             \n",
      "____________________________________________________________________________________________________\n",
      "activation_46 (Activation)       (None, 11)            0           dense_46[0][0]                   \n",
      "____________________________________________________________________________________________________\n",
      "dense_47 (Dense)                 (None, 5)             60          activation_46[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "activation_47 (Activation)       (None, 5)             0           dense_47[0][0]                   \n",
      "____________________________________________________________________________________________________\n",
      "dense_48 (Dense)                 (None, 2)             12          activation_47[0][0]              \n",
      "____________________________________________________________________________________________________\n",
      "activation_48 (Activation)       (None, 2)             0           dense_48[0][0]                   \n",
      "====================================================================================================\n",
      "Total params: 325\n",
      "____________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(11, input_shape=(22,)))\n",
    "model.add(Activation('relu'))\n",
    "\n",
    "model.add(Dense(5))\n",
    "model.add(Activation('relu'))\n",
    "\n",
    "model.add(Dense(2))\n",
    "model.add(Activation('softmax'))\n",
    "\n",
    "model.summary()\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 395,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6499 samples, validate on 1625 samples\n",
      "Epoch 1/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.7276 - acc: 0.5890 - val_loss: 0.7540 - val_acc: 0.4308\n",
      "Epoch 2/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.5810 - acc: 0.6916 - val_loss: 0.6043 - val_acc: 0.6880\n",
      "Epoch 3/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.5132 - acc: 0.7617 - val_loss: 0.4923 - val_acc: 0.8338\n",
      "Epoch 4/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.4594 - acc: 0.8070 - val_loss: 0.4260 - val_acc: 0.8406\n",
      "Epoch 5/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.4140 - acc: 0.8297 - val_loss: 0.3927 - val_acc: 0.8425\n",
      "Epoch 6/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.3762 - acc: 0.8487 - val_loss: 0.3678 - val_acc: 0.8769\n",
      "Epoch 7/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.3423 - acc: 0.8667 - val_loss: 0.3530 - val_acc: 0.8929\n",
      "Epoch 8/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.3138 - acc: 0.8808 - val_loss: 0.3414 - val_acc: 0.8911\n",
      "Epoch 9/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.2841 - acc: 0.8929 - val_loss: 0.3373 - val_acc: 0.8738\n",
      "Epoch 10/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.2520 - acc: 0.9097 - val_loss: 0.3347 - val_acc: 0.8751\n",
      "Epoch 11/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.2169 - acc: 0.9257 - val_loss: 0.3476 - val_acc: 0.8578\n",
      "Epoch 12/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1863 - acc: 0.9368 - val_loss: 0.3444 - val_acc: 0.8578\n",
      "Epoch 13/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1633 - acc: 0.9454 - val_loss: 0.3321 - val_acc: 0.8726\n",
      "Epoch 14/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1445 - acc: 0.9528 - val_loss: 0.3393 - val_acc: 0.8683\n",
      "Epoch 15/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1300 - acc: 0.9594 - val_loss: 0.3327 - val_acc: 0.8751\n",
      "Epoch 16/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1175 - acc: 0.9625 - val_loss: 0.3382 - val_acc: 0.8702\n",
      "Epoch 17/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.1076 - acc: 0.9655 - val_loss: 0.3317 - val_acc: 0.8751\n",
      "Epoch 18/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0999 - acc: 0.9666 - val_loss: 0.3184 - val_acc: 0.8855\n",
      "Epoch 19/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0929 - acc: 0.9695 - val_loss: 0.3326 - val_acc: 0.8745\n",
      "Epoch 20/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0870 - acc: 0.9703 - val_loss: 0.3356 - val_acc: 0.8738\n",
      "Epoch 21/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0820 - acc: 0.9728 - val_loss: 0.3232 - val_acc: 0.8800\n",
      "Epoch 22/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0785 - acc: 0.9738 - val_loss: 0.2950 - val_acc: 0.8966\n",
      "Epoch 23/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0733 - acc: 0.9758 - val_loss: 0.3146 - val_acc: 0.8818\n",
      "Epoch 24/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0695 - acc: 0.9777 - val_loss: 0.3160 - val_acc: 0.8812\n",
      "Epoch 25/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0667 - acc: 0.9783 - val_loss: 0.3126 - val_acc: 0.8831\n",
      "Epoch 26/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0633 - acc: 0.9800 - val_loss: 0.3058 - val_acc: 0.8886\n",
      "Epoch 27/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0608 - acc: 0.9800 - val_loss: 0.3074 - val_acc: 0.8886\n",
      "Epoch 28/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0580 - acc: 0.9815 - val_loss: 0.2915 - val_acc: 0.9022\n",
      "Epoch 29/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0563 - acc: 0.9835 - val_loss: 0.2811 - val_acc: 0.9102\n",
      "Epoch 30/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0569 - acc: 0.9825 - val_loss: 0.3139 - val_acc: 0.8837\n",
      "Epoch 31/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0511 - acc: 0.9849 - val_loss: 0.2683 - val_acc: 0.9138\n",
      "Epoch 32/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0501 - acc: 0.9852 - val_loss: 0.2723 - val_acc: 0.9052\n",
      "Epoch 33/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0474 - acc: 0.9874 - val_loss: 0.2693 - val_acc: 0.9102\n",
      "Epoch 34/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0458 - acc: 0.9874 - val_loss: 0.2454 - val_acc: 0.9194\n",
      "Epoch 35/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0436 - acc: 0.9889 - val_loss: 0.2407 - val_acc: 0.9188\n",
      "Epoch 36/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0425 - acc: 0.9895 - val_loss: 0.2374 - val_acc: 0.9225\n",
      "Epoch 37/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0416 - acc: 0.9902 - val_loss: 0.2502 - val_acc: 0.9194\n",
      "Epoch 38/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0394 - acc: 0.9914 - val_loss: 0.2406 - val_acc: 0.9188\n",
      "Epoch 39/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0377 - acc: 0.9908 - val_loss: 0.2378 - val_acc: 0.9243\n",
      "Epoch 40/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0368 - acc: 0.9912 - val_loss: 0.2374 - val_acc: 0.9218\n",
      "Epoch 41/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0351 - acc: 0.9922 - val_loss: 0.2381 - val_acc: 0.9249\n",
      "Epoch 42/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0339 - acc: 0.9923 - val_loss: 0.2213 - val_acc: 0.9458\n",
      "Epoch 43/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0333 - acc: 0.9923 - val_loss: 0.2271 - val_acc: 0.9360\n",
      "Epoch 44/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0314 - acc: 0.9931 - val_loss: 0.2237 - val_acc: 0.9434\n",
      "Epoch 45/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0312 - acc: 0.9935 - val_loss: 0.2226 - val_acc: 0.9360\n",
      "Epoch 46/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0296 - acc: 0.9943 - val_loss: 0.2222 - val_acc: 0.9397\n",
      "Epoch 47/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0293 - acc: 0.9943 - val_loss: 0.2507 - val_acc: 0.9108\n",
      "Epoch 48/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0281 - acc: 0.9942 - val_loss: 0.2457 - val_acc: 0.9108\n",
      "Epoch 49/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0269 - acc: 0.9948 - val_loss: 0.2497 - val_acc: 0.9095\n",
      "Epoch 50/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0262 - acc: 0.9957 - val_loss: 0.2261 - val_acc: 0.9335\n",
      "Epoch 51/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0254 - acc: 0.9951 - val_loss: 0.2489 - val_acc: 0.9120\n",
      "Epoch 52/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0251 - acc: 0.9946 - val_loss: 0.2381 - val_acc: 0.9175\n",
      "Epoch 53/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0242 - acc: 0.9955 - val_loss: 0.2345 - val_acc: 0.9182\n",
      "Epoch 54/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0233 - acc: 0.9960 - val_loss: 0.2515 - val_acc: 0.9003\n",
      "Epoch 55/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0228 - acc: 0.9965 - val_loss: 0.2293 - val_acc: 0.9200\n",
      "Epoch 56/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0216 - acc: 0.9963 - val_loss: 0.2175 - val_acc: 0.9366\n",
      "Epoch 57/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0213 - acc: 0.9963 - val_loss: 0.2137 - val_acc: 0.9323\n",
      "Epoch 58/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0200 - acc: 0.9965 - val_loss: 0.2302 - val_acc: 0.9120\n",
      "Epoch 59/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0198 - acc: 0.9963 - val_loss: 0.2253 - val_acc: 0.9169\n",
      "Epoch 60/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0188 - acc: 0.9972 - val_loss: 0.2182 - val_acc: 0.9225\n",
      "Epoch 61/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0182 - acc: 0.9974 - val_loss: 0.1971 - val_acc: 0.9385\n",
      "Epoch 62/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0179 - acc: 0.9969 - val_loss: 0.1963 - val_acc: 0.9342\n",
      "Epoch 63/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0171 - acc: 0.9968 - val_loss: 0.2839 - val_acc: 0.8775\n",
      "Epoch 64/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0165 - acc: 0.9971 - val_loss: 0.1792 - val_acc: 0.9428\n",
      "Epoch 65/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0155 - acc: 0.9977 - val_loss: 0.2325 - val_acc: 0.9089\n",
      "Epoch 66/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0159 - acc: 0.9969 - val_loss: 0.1963 - val_acc: 0.9298\n",
      "Epoch 67/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0145 - acc: 0.9975 - val_loss: 0.1578 - val_acc: 0.9569\n",
      "Epoch 68/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0139 - acc: 0.9977 - val_loss: 0.1967 - val_acc: 0.9298\n",
      "Epoch 69/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0133 - acc: 0.9977 - val_loss: 0.1997 - val_acc: 0.9292\n",
      "Epoch 70/70\n",
      "6499/6499 [==============================] - 0s - loss: 0.0124 - acc: 0.9977 - val_loss: 0.2114 - val_acc: 0.9237\n"
     ]
    }
   ],
   "source": [
    "hist = model.fit(x_train, Y_train,\n",
    "         nb_epoch=70,\n",
    "         shuffle=True,\n",
    "         batch_size=128,\n",
    "         validation_data=(x_test, Y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 396,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJAAAAEoCAYAAAAHcQuWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCiCJiIiIiIiIiEiR/j9ESCpgwqKErQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1117483d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14,3))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.suptitle('Optimizer : Adam', fontsize=10)\n",
    "plt.ylabel('Loss', fontsize=16)\n",
    "plt.plot(hist.history['loss'], 'b', label='Training Loss')\n",
    "plt.plot(hist.history['val_loss'], 'r', label='Validation Loss')\n",
    "plt.legend(loc='upper right')\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.ylabel('Accuracy', fontsize=16)\n",
    "plt.plot(hist.history['acc'], 'b', label='Training Accuracy')\n",
    "plt.plot(hist.history['val_acc'], 'r', label='Validation Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Plots with different Optimizers are ploted below ##"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![alt text](https://cloud.githubusercontent.com/assets/15228557/21239673/ebc262ac-c32d-11e6-80da-08e35b4fde27.jpg \"Plot with RMSprop Optimizer\")\n",
    "\n",
    "<br><br>\n",
    "\n",
    "![alt text](https://cloud.githubusercontent.com/assets/15228557/21239677/f45d369e-c32d-11e6-940a-f019fd82a283.jpg \"Plot with Adagrad Optimizer\")\n",
    "\n",
    "<br><br>\n",
    "\n",
    "![alt text](https://cloud.githubusercontent.com/assets/15228557/21239685/fb7c3574-c32d-11e6-87dc-47bcecd79b94.jpg \"Plot with Stochastic Gradient Descent\")\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}