Recognizing-CIFAR-10-images-Improved-Model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Recognizing CIFAR-10 images with Deep Learning (Part II)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "CIFAR-10 dataset contains 60,000 color images of size 32*32 pixels which are divided in 10 classes.\n",
    "We will use Keras library to import the dataset which contains 50,000 images as training iamges\n",
    "and 10,000 as test images\n",
    "\n",
    "In [Part I](https://github.com/chhayac/Machine-Learning-Notebooks/blob/master/Recognizing-CIFAR-10-images-Simple-Model.ipynb), we used a simple Convolutional Neural Network model with only one hidden layer. Model gave accuracy of 66.94% on test data. Goal of this part is to improve the accuracy of the model by increasing the number of convolutional layers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from keras.datasets import cifar10\n",
    "from keras.utils import np_utils\n",
    "from keras.optimizers import SGD, Adam, RMSprop\n",
    "from keras.models import Sequential\n",
    "from keras.layers.convolutional import Conv2D, MaxPooling2D\n",
    "from keras.layers.core import Dense, Flatten, Activation, Dropout\n",
    "import matplotlib.pyplot as plt\n",
    "% matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "#load dataset\n",
    "(X_train, y_train), (X_test, y_test) = cifar10.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "BATCH_SIZE = 128\n",
    "nb_epochs = 40\n",
    "VALIDATION_SPLIT = 0.2\n",
    "num_classes = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_train, img_channels, img_rows, img_cols = X_train.shape\n",
    "num_test, _, _, _= X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train Shape: (50000, 32, 32, 3)\n",
      "Training samples: 50000\n",
      "Test Samples: 10000\n"
     ]
    }
   ],
   "source": [
    "print('X_train Shape:', X_train.shape)\n",
    "training_samples = X_train.shape[0]\n",
    "print(\"Training samples:\", training_samples)\n",
    "print(\"Test Samples:\", y_test.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names =['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1342f4be3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize =(8, 3))\n",
    "for i in range(num_classes):\n",
    "    axis = fig.add_subplot(2, 5, 1+i, xticks = [], yticks = [])\n",
    "    idx = np.where(y_train[:] == i)[0]\n",
    "    features_idx = X_train[idx, ::]\n",
    "    image_num = np.random.randint(features_idx.shape[0])\n",
    "    image = features_idx[image_num,::]\n",
    "    axis.set_title(class_names[i])\n",
    "    plt.imshow(image)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One hot Encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert to categorical\n",
    "y_train = np_utils.to_categorical(y_train, num_classes)\n",
    "y_test = np_utils.to_categorical(y_test, num_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "#float and normalization\n",
    "X_train = X_train.astype('float32')\n",
    "X_test = X_test.astype('float32')\n",
    "X_train /= 255\n",
    "X_test /= 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "if K.backend() == 'tensorflow':\n",
    "    K.set_image_dim_ordering('tf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Building sequential model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Network\n",
    "model = Sequential()\n",
    "model.add(Conv2D(32, (3, 3), padding = 'same', input_shape=X_train.shape[1:]))\n",
    "model.add(Activation('relu'))\n",
    "\n",
    "# Adding more layers to improve the model\n",
    "model.add(Conv2D(32, (3, 3), padding = 'same'))\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size= (2,2)))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Conv2D(64, (3, 3), padding = 'same'))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Conv2D(64, (3, 3)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size= (2, 2)))\n",
    "model.add(Dropout(0.25))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_50 (Conv2D)           (None, 32, 32, 32)        896       \n",
      "_________________________________________________________________\n",
      "activation_43 (Activation)   (None, 32, 32, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_51 (Conv2D)           (None, 32, 32, 32)        9248      \n",
      "_________________________________________________________________\n",
      "activation_44 (Activation)   (None, 32, 32, 32)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_17 (MaxPooling (None, 16, 16, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_17 (Dropout)         (None, 16, 16, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_52 (Conv2D)           (None, 16, 16, 64)        18496     \n",
      "_________________________________________________________________\n",
      "activation_45 (Activation)   (None, 16, 16, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_53 (Conv2D)           (None, 14, 14, 64)        36928     \n",
      "_________________________________________________________________\n",
      "activation_46 (Activation)   (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_18 (MaxPooling (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_18 (Dropout)         (None, 7, 7, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 3136)              0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 512)               1606144   \n",
      "_________________________________________________________________\n",
      "activation_47 (Activation)   (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dropout_19 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 10)                5130      \n",
      "_________________________________________________________________\n",
      "activation_48 (Activation)   (None, 10)                0         \n",
      "=================================================================\n",
      "Total params: 1,676,842\n",
      "Trainable params: 1,676,842\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.add(Flatten())\n",
    "model.add(Dense(512))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes))\n",
    "model.add(Activation('softmax'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compiling the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss= 'categorical_crossentropy', optimizer = 'rmsprop', metrics= ['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 40000 samples, validate on 10000 samples\n",
      "Epoch 1/40\n",
      "40000/40000 [==============================] - 18s 440us/step - loss: 1.5278 - acc: 0.4518 - val_loss: 1.6131 - val_acc: 0.4525\n",
      "Epoch 2/40\n",
      "40000/40000 [==============================] - 17s 419us/step - loss: 1.2531 - acc: 0.5565 - val_loss: 1.1684 - val_acc: 0.5918\n",
      "Epoch 3/40\n",
      "40000/40000 [==============================] - 17s 417us/step - loss: 1.0798 - acc: 0.6193 - val_loss: 1.0287 - val_acc: 0.6417\n",
      "Epoch 4/40\n",
      "40000/40000 [==============================] - 16s 409us/step - loss: 0.9607 - acc: 0.6632 - val_loss: 0.9146 - val_acc: 0.6791\n",
      "Epoch 5/40\n",
      "40000/40000 [==============================] - 16s 408us/step - loss: 0.8759 - acc: 0.6944 - val_loss: 0.8480 - val_acc: 0.6947\n",
      "Epoch 6/40\n",
      "40000/40000 [==============================] - 17s 425us/step - loss: 0.8091 - acc: 0.7160 - val_loss: 0.8278 - val_acc: 0.7159\n",
      "Epoch 7/40\n",
      "40000/40000 [==============================] - 18s 447us/step - loss: 0.7535 - acc: 0.7385 - val_loss: 0.7752 - val_acc: 0.7303\n",
      "Epoch 8/40\n",
      "40000/40000 [==============================] - 17s 415us/step - loss: 0.7058 - acc: 0.7541 - val_loss: 0.7922 - val_acc: 0.7342\n",
      "Epoch 9/40\n",
      "40000/40000 [==============================] - 17s 435us/step - loss: 0.6639 - acc: 0.7700 - val_loss: 0.8233 - val_acc: 0.7289\n",
      "Epoch 10/40\n",
      "40000/40000 [==============================] - 17s 432us/step - loss: 0.6287 - acc: 0.7824 - val_loss: 0.7478 - val_acc: 0.7593\n",
      "Epoch 11/40\n",
      "40000/40000 [==============================] - 16s 409us/step - loss: 0.6053 - acc: 0.7906 - val_loss: 0.8659 - val_acc: 0.7143\n",
      "Epoch 12/40\n",
      "40000/40000 [==============================] - 17s 419us/step - loss: 0.5816 - acc: 0.7991 - val_loss: 0.7777 - val_acc: 0.7617\n",
      "Epoch 13/40\n",
      "40000/40000 [==============================] - 17s 425us/step - loss: 0.5621 - acc: 0.8047 - val_loss: 0.7376 - val_acc: 0.7693\n",
      "Epoch 14/40\n",
      "40000/40000 [==============================] - 18s 438us/step - loss: 0.5505 - acc: 0.8088 - val_loss: 0.7019 - val_acc: 0.7741\n",
      "Epoch 15/40\n",
      "40000/40000 [==============================] - 17s 432us/step - loss: 0.5442 - acc: 0.8141 - val_loss: 0.7530 - val_acc: 0.7661\n",
      "Epoch 16/40\n",
      "40000/40000 [==============================] - 17s 430us/step - loss: 0.5349 - acc: 0.8198 - val_loss: 1.0331 - val_acc: 0.7229\n",
      "Epoch 17/40\n",
      "40000/40000 [==============================] - 17s 429us/step - loss: 0.5285 - acc: 0.8203 - val_loss: 0.7996 - val_acc: 0.7374\n",
      "Epoch 18/40\n",
      "40000/40000 [==============================] - 17s 429us/step - loss: 0.5260 - acc: 0.8206 - val_loss: 0.7752 - val_acc: 0.7723\n",
      "Epoch 19/40\n",
      "40000/40000 [==============================] - 17s 431us/step - loss: 0.5099 - acc: 0.8291 - val_loss: 0.6853 - val_acc: 0.7755\n",
      "Epoch 20/40\n",
      "40000/40000 [==============================] - 17s 434us/step - loss: 0.5194 - acc: 0.8254 - val_loss: 0.7063 - val_acc: 0.7870\n",
      "Epoch 21/40\n",
      "40000/40000 [==============================] - 17s 437us/step - loss: 0.5126 - acc: 0.8294 - val_loss: 0.8614 - val_acc: 0.7751\n",
      "Epoch 22/40\n",
      "40000/40000 [==============================] - 17s 436us/step - loss: 0.5255 - acc: 0.8282 - val_loss: 0.7333 - val_acc: 0.7806\n",
      "Epoch 23/40\n",
      "40000/40000 [==============================] - 17s 433us/step - loss: 0.5146 - acc: 0.8298 - val_loss: 1.0159 - val_acc: 0.7522\n",
      "Epoch 24/40\n",
      "40000/40000 [==============================] - 17s 432us/step - loss: 0.5155 - acc: 0.8287 - val_loss: 0.7216 - val_acc: 0.7811\n",
      "Epoch 25/40\n",
      "40000/40000 [==============================] - 16s 408us/step - loss: 0.5154 - acc: 0.8301 - val_loss: 0.8039 - val_acc: 0.7872\n",
      "Epoch 26/40\n",
      "40000/40000 [==============================] - 16s 405us/step - loss: 0.5114 - acc: 0.8317 - val_loss: 0.7967 - val_acc: 0.7842\n",
      "Epoch 27/40\n",
      "40000/40000 [==============================] - 16s 405us/step - loss: 0.5111 - acc: 0.8339 - val_loss: 1.1211 - val_acc: 0.7714\n",
      "Epoch 28/40\n",
      "40000/40000 [==============================] - 18s 441us/step - loss: 0.5223 - acc: 0.8309 - val_loss: 0.7575 - val_acc: 0.7831\n",
      "Epoch 29/40\n",
      "40000/40000 [==============================] - 17s 437us/step - loss: 0.5242 - acc: 0.8290 - val_loss: 0.6573 - val_acc: 0.7826\n",
      "Epoch 30/40\n",
      "40000/40000 [==============================] - 17s 434us/step - loss: 0.5205 - acc: 0.8280 - val_loss: 0.6888 - val_acc: 0.7854\n",
      "Epoch 31/40\n",
      "40000/40000 [==============================] - 17s 430us/step - loss: 0.5293 - acc: 0.8289 - val_loss: 0.9044 - val_acc: 0.7785\n",
      "Epoch 32/40\n",
      "40000/40000 [==============================] - 17s 427us/step - loss: 0.5207 - acc: 0.8296 - val_loss: 0.8126 - val_acc: 0.7892\n",
      "Epoch 33/40\n",
      "40000/40000 [==============================] - 17s 434us/step - loss: 0.5242 - acc: 0.8280 - val_loss: 0.7669 - val_acc: 0.7763\n",
      "Epoch 34/40\n",
      "40000/40000 [==============================] - 17s 424us/step - loss: 0.5292 - acc: 0.8287 - val_loss: 0.7010 - val_acc: 0.7820\n",
      "Epoch 35/40\n",
      "40000/40000 [==============================] - 17s 427us/step - loss: 0.5294 - acc: 0.8293 - val_loss: 0.8707 - val_acc: 0.7793\n",
      "Epoch 36/40\n",
      "40000/40000 [==============================] - 17s 428us/step - loss: 0.5285 - acc: 0.8289 - val_loss: 0.7430 - val_acc: 0.7592\n",
      "Epoch 37/40\n",
      "40000/40000 [==============================] - 17s 425us/step - loss: 0.5380 - acc: 0.8281 - val_loss: 0.8493 - val_acc: 0.7674\n",
      "Epoch 38/40\n",
      "40000/40000 [==============================] - 17s 432us/step - loss: 0.5371 - acc: 0.8287 - val_loss: 0.8002 - val_acc: 0.7833\n",
      "Epoch 39/40\n",
      "40000/40000 [==============================] - 17s 429us/step - loss: 0.5402 - acc: 0.8271 - val_loss: 0.8141 - val_acc: 0.7809\n",
      "Epoch 40/40\n",
      "40000/40000 [==============================] - 17s 423us/step - loss: 0.5442 - acc: 0.8252 - val_loss: 0.8364 - val_acc: 0.7947\n"
     ]
    }
   ],
   "source": [
    "model_history = model.fit(X_train, y_train, batch_size = BATCH_SIZE, epochs = nb_epochs, \n",
    "          validation_split = VALIDATION_SPLIT, verbose = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "List all data in history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['loss', 'val_loss', 'val_acc', 'acc'])\n"
     ]
    }
   ],
   "source": [
    "print(model_history.history.keys())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Summarize history for accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x133a032e828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model_history.history['acc'])\n",
    "plt.plot(model_history.history['val_acc'])\n",
    "plt.title(\"Model Accuracy\")\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(['train','validation'], loc = 'upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Summarize history for loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x133a268b198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model_history.history['loss'])\n",
    "plt.plot(model_history.history['val_loss'])\n",
    "plt.title(\"Model loss\")\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.legend(['train','validation'], loc = 'upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 2s 179us/step\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(X_test, y_test, batch_size = BATCH_SIZE, verbose =1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score 0.880404548073\n",
      "Test accuracy 0.7865\n"
     ]
    }
   ],
   "source": [
    "print(\"Test Score\", score[0])\n",
    "print(\"Test accuracy\", score[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This accuracy is 12% better than what we got in Part I."
   ]
  }
 ],
 "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": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}