mnist tutorial with keras

mnist_cnn.py

# https://www.tensorflow.org/tutorials/images/cnn

from __future__ import absolute_import, division, print_function, unicode_literals


import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout, Conv2D, MaxPooling2D, Flatten
from keras.optimizers import RMSprop


(train_images, train_labels), (test_images, test_labels) = mnist.load_data()

train_images = train_images.reshape((train_images.shape[0],
                                     train_images.shape[1],
                                     train_images.shape[2],
                                     1))
test_images = test_images.reshape((test_images.shape[0],
                                   test_images.shape[1],
                                   test_images.shape[2],
                                   1))

train_images, test_images = train_images / 255.0, test_images / 255.0

model = Sequential()
# 32 output channels, 3 * 3 filter
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=train_images.shape[1:]))
# 2 * 2 pool size
model.add(MaxPooling2D((2, 2)))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D((2, 2)))
model.add(Flatten())
model.add(Dense(64, activation='relu'))
model.add(Dense(10, activation='softmax'))

model.summary()

model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.fit(train_images, train_labels, epochs=5)


score = model.evaluate(test_images, test_labels, verbose=2)

print('Test loss:', score[0])
print('Test accuracy:', score[1])

mnist_mlp.py

# Shakyou: https://github.com/keras-team/keras/blob/master/examples/mnist_mlp.py
# Kaisetsu: https://qiita.com/ash8h/items/29e24fc617b832fba136#%E3%83%A2%E3%83%87%E3%83%AB%E6%A7%8B%E7%AF%89


from __future__ import print_function


import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout
from keras.optimizers import RMSprop


batch_size = 128
num_classes = 10
epochs = 20


(in_train, out_train), (in_test, out_test) = mnist.load_data()


# shape: (num_data, number of rows, number of columns)
in_train = in_train.reshape(in_train.shape[0], in_train.shape[1] * in_train.shape[2])
in_train = in_train.astype('float32')
in_train /= 255

in_test = in_test.reshape(in_test.shape[0], in_test.shape[1] * in_test.shape[2])
in_test = in_test.astype('float32')
in_test /= 255


out_train = keras.utils.to_categorical(out_train, num_classes)
out_test = keras.utils.to_categorical(out_test, num_classes)

model = Sequential()
model.add(Dense(512, activation='relu', input_shape=in_train.shape[1:]))
model.add(Dropout(0.2))
model.add(Dense(512, activation='relu'))
model.add(Dropout(0.2))
model.add(Dense(num_classes, activation='softmax'))


model.summary()


model.compile(RMSprop(), 'categorical_crossentropy', ['accuracy'])

history = model.fit(in_train, out_train,
                    batch_size=batch_size,
                    epochs=epochs,
                    verbose=1,
                    validation_data=(in_test, out_test))
score = model.evaluate(in_test, out_test, verbose=0)

print('Test loss:', score[0])
print('Test accuracy:', score[1])