Initialize dependencies and get data

In [1]:
import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D
from keras import backend as K

batch_size = 128
num_classes = 10
epochs = 12

# input image dimensions
img_rows, img_cols = 28, 28

# the data, shuffled and split between train and test sets
(x_train, y_train), (x_test, y_test) = mnist.load_data()

if K.image_data_format() == 'channels_first':
    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)
    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)
    input_shape = (1, img_rows, img_cols)
else:
    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)
    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)
    input_shape = (img_rows, img_cols, 1)
Using TensorFlow backend.
Downloading data from https://s3.amazonaws.com/img-datasets/mnist.npz
11444224/11490434 [============================>.] - ETA: 0s

Pre-process image data

In [3]:
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
print('x_train shape:', x_train.shape)
print(x_train.shape[0], 'train samples')
print(x_test.shape[0], 'test samples')
x_train shape: (60000, 28, 28, 1)
60000 train samples
10000 test samples
In [5]:
# convert class vectors to binary class matrices
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)

Build a CNN based deep neural network

In [6]:
model = Sequential()
model.add(Conv2D(32, kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(num_classes, activation='softmax'))

Visualize the network architecture

In [7]:
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot

SVG(model_to_dot(model, show_shapes=True, 
                 show_layer_names=True, rankdir='TB').create(prog='dot', format='svg'))
Out[7]:
G 1648868526512 conv2d_1_input: InputLayer input: output: (None, 28, 28, 1) (None, 28, 28, 1) 1648868526176 conv2d_1: Conv2D input: output: (None, 28, 28, 1) (None, 26, 26, 32) 1648868526512->1648868526176 1648868702080 conv2d_2: Conv2D input: output: (None, 26, 26, 32) (None, 24, 24, 64) 1648868526176->1648868702080 1648868701968 max_pooling2d_1: MaxPooling2D input: output: (None, 24, 24, 64) (None, 12, 12, 64) 1648868702080->1648868701968 1648868661288 dropout_1: Dropout input: output: (None, 12, 12, 64) (None, 12, 12, 64) 1648868701968->1648868661288 1648849747640 flatten_1: Flatten input: output: (None, 12, 12, 64) (None, 9216) 1648868661288->1648849747640 1648849891904 dense_1: Dense input: output: (None, 9216) (None, 128) 1648849747640->1648849891904 1648849845104 dropout_2: Dropout input: output: (None, 128) (None, 128) 1648849891904->1648849845104 1648849975784 dense_2: Dense input: output: (None, 128) (None, 10) 1648849845104->1648849975784

Build the model

In [8]:
model.compile(loss=keras.losses.categorical_crossentropy,
              optimizer=keras.optimizers.Adadelta(),
              metrics=['accuracy'])
In [10]:
model.fit(x_train, y_train,
          batch_size=batch_size,
          epochs=2,
          verbose=1,
          validation_data=(x_test, y_test))
Train on 60000 samples, validate on 10000 samples
Epoch 1/2
60000/60000 [==============================] - 230s - loss: 0.1867 - acc: 0.9444 - val_loss: 0.0692 - val_acc: 0.9786
Epoch 2/2
60000/60000 [==============================] - 232s - loss: 0.1061 - acc: 0.9682 - val_loss: 0.0524 - val_acc: 0.9835
Out[10]:
<keras.callbacks.History at 0x17fe7308240>

Predict and test model performance

In [11]:
score = model.evaluate(x_test, y_test, verbose=1)
10000/10000 [==============================] - 13s    
In [12]:
print('Test loss:', score[0])
print('Test accuracy:', score[1])
Test loss: 0.0523672130647
Test accuracy: 0.9835