Keras RNN Cyclic neural network (IV.)

To import the desired lib:

From keras.datasets import  mnist to
keras.utils import np_utils from
keras.models import sequential
From keras.layers import dense,dropout,activation,simplernn from
keras.optimizers import  Adam
Import NumPy as NP

To set a random seed:

Np.random.seed (1337)

Set some parameters:

time_steps=28  #时间点, length of time (28 lines)
input_size=28  #每一行 Read how many 
ouput_size=10  #输出 results one_hot
Cell_size=50   # How many unit lr=0.001 to put on a hidden layer in the network       #学习率

Load data:

(X_train,y_train), (x_test,y_test) =mnist.load_data () #加载数据

Data preprocessing:

#数据预处理
X_train=x_train.reshape ( -1,28,28)/255 
x_test=x_test.reshape ( -1,28,28)/255 
y_train=np_ Utils.to_categorical (y_train,num_classes=10)
y_test=np_utils.to_categorical (y_test,num_classes=10)

To create a model:

#建型
model=sequential ()

RNN:

#RNN
model.add (simplernn
        cell_size,
        input_shape= (time_steps,input_size), # or:input_dim=input_size, Input_length=time_steps,

))

Output layer:

Model.add (Dense (ouput_size))
Model.add (Activation (' Softmax '))

Optimizer:

#优化器
Adam=adam (LR)

Parameter information for the output model:

Model.summary ()

Build the Model:

#搭建模型
model.compile (optimizer= ' Adam ', loss= ' categorical_crossentropy ', metrics=[' accuracy '])

Training:

Model.fit (X_TRAIN,Y_TRAIN,EPOCHS=5,BATCH_SIZE=32)

Evaluation:

Loss,accuracy=model.evaluate (x_test,y_test)
print (loss)
print (accuracy)

Results:

9888/10000 [============================>]-eta:0s0.188649062154
0.9443