我在double for循环中使用相同的keras模型,但要确保迭代完全独立,如下所示。
Dense network
LSTM网络
X = concatenate([Dense,LSTM])
Y = Dense(1)(X)
model = Model(inputs=[Dense_input,LSTMN_input], outputs=Y)
model.compile(loss='mse', optimizer='RMSProp', metrics=['accuracy'])
es = EarlyStopping(monitor='val_acc', min_delta=0, patience=10, restore_best_weights=True)
for q in range(A):
```
add new samples to training data
```
for w in range(B):
```
model.fit([dense, lstm],output)
```
```
make some predictions and get new samples
实现此目的的有效方法是什么?