import numpy as np
from keras.datasets import mnist
from keras.utils import np_utils
from keras.models import Sequential
from keras.layers import Dense,Dropout,Convolution2D,MaxPooling2D,Flatten  # Flatten扁平化
from keras.optimizers import Adam

# 载入数据
(x_train,y_train),(x_test,y_test) = mnist.load_data()
# (60000,28,28)->(60000,28,28,1)
x_train = x_train.reshape(-1,28,28,1)/255.0  # 归一化
x_test = x_test.reshape(-1,28,28,1)/255.0
# 换one hot格式
y_train = np_utils.to_categorical(y_train,num_classes=10)
y_test = np_utils.to_categorical(y_test,num_classes=10)

# 定义顺序模型
model = Sequential()

# 第一个卷积层
# input_shape 输入平面
# filters 卷积核/滤波器个数
# kernel_size 卷积窗口大小
# strides 步长
# padding padding方式 same/valid
# activation 激活函数
model.add(Convolution2D(
    input_shape = (28,28,1),
    filters = 32,
    kernel_size = 5,
    strides = 1,
    padding = 'same',
    activation = 'relu'
))
# 第一个池化层  相当于缩小一倍
model.add(MaxPooling2D(
    pool_size = 2,
    strides = 2,
    padding = 'same',
))
# 第二个卷积层
model.add(Convolution2D(64,5,strides=1,padding='same',activation = 'relu'))
# 第二个池化层
model.add(MaxPooling2D(2,2,'same'))
# 把第二个池化层的输出扁平化为1维
model.add(Flatten())
# 第一个全连接层
model.add(Dense(1024,activation = 'relu'))
# Dropout
model.add(Dropout(0.5))
# 第二个全连接层
model.add(Dense(10,activation='softmax'))

# 定义优化器
adam = Adam(lr=1e-4)  # 10的负4次方

# 定义优化器，loss function，训练过程中计算准确率
model.compile(optimizer=adam,loss='categorical_crossentropy',metrics=['accuracy'])

# 训练模型
model.fit(x_train,y_train,batch_size=64,epochs=10)

# 评估模型
loss,accuracy = model.evaluate(x_test,y_test)

print('test loss',loss)
print('test accuracy',accuracy)

结果

Epoch 1/10
60000/60000 [==============================] - 21s 345us/step - loss: 0.3263 - acc: 0.9087
Epoch 2/10
60000/60000 [==============================] - 15s 247us/step - loss: 0.0912 - acc: 0.9722
Epoch 3/10
60000/60000 [==============================] - 15s 248us/step - loss: 0.0631 - acc: 0.98120s - loss: 0.06
Epoch 4/10
60000/60000 [==============================] - 15s 248us/step - loss: 0.0489 - acc: 0.98490s - loss: 0.048
Epoch 5/10
60000/60000 [==============================] - 15s 248us/step - loss: 0.0411 - acc: 0.9872
Epoch 6/10
60000/60000 [==============================] - 15s 257us/step - loss: 0.0345 - acc: 0.9899
Epoch 7/10
60000/60000 [==============================] - 15s 248us/step - loss: 0.0299 - acc: 0.9902
Epoch 8/10
60000/60000 [==============================] - 15s 257us/step - loss: 0.0253 - acc: 0.9920
Epoch 9/10
60000/60000 [==============================] - 16s 263us/step - loss: 0.0226 - acc: 0.9929
Epoch 10/10
60000/60000 [==============================] - 15s 252us/step - loss: 0.0200 - acc: 0.9936
10000/10000 [==============================] - 1s 113us/step
test loss 0.020262729431517072
test accuracy 0.9928