04 如何构建一个最简单的卷积神经网络

import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D
from keras.utils import plot_model

# 1.超参数
batch_size = 128
num_classes = 10
epochs = 12
img_rows, img_cols = 28, 28

# 2.加载数据
(x_train, y_train), (x_test, y_test) = mnist.load_data()
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)
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)

# 3.搭建模型
model = Sequential()
# 卷积时候选择same:W/S VALID:(W-F+1)/S 均向下取整 (28-2+1)/2
model.add(Conv2D(filters=32,
                 kernel_size=(3, 3),
                 activation='relu',
                 input_shape=input_shape))
model.add(Conv2D(filters=64,
                 kernel_size=(3, 3),
                 activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(rate=0.25))
# 连接全连接层的时候需要展平
model.add(Flatten())
model.add(Dropout(rate=0.5))
# 最后接softmax，
model.add(Dense(units=num_classes, activation='softmax'))

# 4. 编译模型
model.compile(loss='categorical_crossentropy',
              optimizer='Adadelta',
              metrics=['accuracy'])

# 5. 训练模型
model.fit(x=x_train, y=y_train, batch_size=batch_size,
          epochs=epochs, verbose=1, validation_data=(x_test, y_test))

# 6.评估模型
socre = model.evaluate(x=x_test, y=y_test, verbose=0)
plot_model(model=model, to_file='models/01mnist_cnn.png', show_shapes=True)