X_test = X_test.astype('float32') / 255
# Convert class vectors to binary class matrices.
Y_train = keras.utils.to_categorical(Y_train, NB_CLASSES)
Y_test = keras.utils.to_categorical(Y_test, NB_CLASSES)
'''人为削减'''
X_train = X_train[0:1000]
Y_train = Y_train[0:1000]
X_test = X_test[0:200]
Y_test = Y_test[0:200]
# initiate RMSprop optimizer
opt = keras.optimizers.rmsprop(lr=0.001, decay=1e-6)
# train the model using RMSprop
model = CNN.createCNN(INPUT_SHAPE, NB_CLASSES)
model.compile(loss='categorical_crossentropy',
optimizer=opt,
metrics=['accuracy'])
modelfile = 'modelweight_10percent.model' #神经网络权重保存
file_path_history = 'historyfile.bin' #保存history，留着作图
if os.path.exists(modelfile): #如果存在之前训练的权重矩阵，载入模型
print('载入模型参数')
model.load_weights(modelfile)
else:
print('训练')
history = model.fit(X_train,
Y_train,
batch_size=BATCH_SIZE,
