目的：实现mnist分类1.加载mnist图像数据(train_x,train_y),(test_x,test_y)=mnist.load_data()2.修改数据类型，数据归一化#将数据转为浮点数x_train=train_x。astype('float32')y_train=train_y.astype('float32')x_test=test_x.astype('float32')y_test=test_y.astype('float32')#数据归一化x_train/=255x_test/=255#标签one_hont编码y_train=keras.utils.to_categorical(y_train,num_classes=10)y_test=keras.utils.to_categorical(y_test,num_classes=10)修改数据形状x_train=x_train.reshape(60000,28,28,1)x_test=x_test.reshape(10000,28,28,1)4.构建学习网络模型=Sequential([Conv2D(filters=6,kernel_size=(5,5),padding='valid',activation='tanh',input_shape=(28,28,1)),MaxPooling2D(pool_size=(2,2)),Conv2D(filters=16,kernel_size=(5,5),padding='valid',activation='tanh'),MaxPooling2D(pool_size=(2,2)),Flatten(),Dense(120,activation='tanh'),Dense(84,activation='tanh'),密集(10,activation='softmax')])5.配置学习网络model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['acc'])6.开始训练model.fit(x_train,y_train,batch_size=64,epochs=20,verbose=1,validation_split=0.2,shuffle=True)7.Evaluationnetworkscore=model.evaluate(x_test,y_test,batch_size=64)print(score)遇到问题：1.ValueError:Negativedimensionsizecausedbysubtracting5from1用于“conv2d_1/卷积”（操作：“Conv2D”），输入形状：[?,1,28,28]，[5,5,28,6]。这是因为原案例使用Theano作为keras的背景，Theano的格式是(1,28,28)。如果是TensorFlow，需要修改inputformat为(28,28,1)2.检查输入时报错：expectedconv2d_1_inputtohave4dimensions,butgotarraywithshape(60000,28,28)直接导入的mnist数据是(60000,28,28)，而你的cnn网络需要的是(60000,28,28,1)，所以你需要修改矩阵形状为(60000,28,28,1)x_train=x_train.reshape(60000,28,28,1)x_test=x_test.reshape(10000,28,28,1)完整代码importkerasfromkeras.datasetsimportmnistfromkeras.modelsimportSequentialfromkeras.layersimportDense,Flattenfromkeras.layersimportConv2D，最大Pooling2Dif__name__==“__main__”:(train_x,train_y),(test_x,test_y)=mnist.load_data()x_train=train_x.astype('float32')y_train=train_y.astype('float32')x_test=test_x.astype('float32')y_test=test_y.astype('float32')x_train/=255x_test/=255y_train=keras.utils.to_categorical(y_train,num_classes=10)y_test=keras.utils.to_categorical(y_test,num_classes=10)x_train=x_train.reshape(60000,28,28,1)x_test=x_test.reshape(10000,28,28,1)model=Sequential([Conv2D(filters=6,kernel_size=(5,5),padding='有效',activation='relu',input_shape=(28,28,1)),MaxPooling2D(pool_size=(2,2)),Conv2D(filters=16,kernel_size=(5,5),padding='valid',activation='relu'),MaxPooling2D(pool_size=(2,2)),Flatten(),Dense(120,activation='relu'),Dense(84,activation='relu'),Dense(10,activation='softmax')])model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['acc'])model.fit(x_train,y_train,batch_size=64,epochs=20,verbose=1,validation_split=0.2,shuffle=True)score=model。评估（x_test，y_test，batch_size=64）打印（分数）model.save（'lenet5.h5'）
