ch5-1.py

import os
import sys

try:
    base_directory = os.path.split(sys.executable)[0]
    os.environ['PATH'] += ';' + base_directory
    import cntk
    os.environ['KERAS_BACKEND'] = 'cntk'
except ImportError:
    print('CNTK not installed')

import keras
import keras.utils
import keras.datasets
import keras.models
import keras.layers
import numpy as np
import matplotlib.pyplot as plt


(train_images, train_labels), (test_images, test_labels) = keras.datasets.mnist.load_data()
train_images = train_images.reshape((60000, 28, 28, 1))
train_images = train_images.astype('float32') / 255
test_images = test_images.reshape((10000, 28, 28, 1))
test_images = test_images.astype('float32') / 255
train_labels = keras.utils.to_categorical(train_labels)
test_labels = keras.utils.to_categorical(test_labels)

model = keras.models.Sequential()
model.add(keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)))
model.add(keras.layers.MaxPooling2D((2, 2)))
model.add(keras.layers.Conv2D(64, (3, 3), activation='relu'))
model.add(keras.layers.MaxPooling2D((2, 2)))
model.add(keras.layers.Conv2D(64, (3, 3), activation='relu'))
model.add(keras.layers.Flatten())
model.add(keras.layers.Dense(64, activation='relu'))
model.add(keras.layers.Dense(10, activation='softmax'))
model.summary()

model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])
model.fit(train_images, train_labels, epochs=5, batch_size=64)

test_loss, test_acc = model.evaluate(test_images, test_labels)
print('test_acc={0:.4f}'.format(test_acc))