如何在LSTM中有效使用批量归一化？

Question

我正在尝试使用 R 中的 keras 在 LSTM 中使用批量归一化。在我的数据集中，目标/输出变量是列Sales，数据集中的每一行记录Sales一年中的每一天（2008-2017）。数据集如下所示：

我的目标是基于这样的数据集构建一个 LSTM 模型，该模型应该能够在训练结束时提供预测。我正在使用 2008-2016 年的数据训练这个模型，并使用 2017 年数据的一半作为验证，其余作为测试集。

之前，我尝试使用 dropout 和提前停止创建模型。如下所示：

mdl1 <- keras_model_sequential()
mdl1 %>%
  layer_lstm(units = 512, input_shape = c(1, 3), return_sequences = T ) %>%  
  layer_dropout(rate = 0.3) %>%
  layer_lstm(units = 512, return_sequences = FALSE) %>%
  layer_dropout(rate = 0.2) %>%
  layer_dense(units = 1, activation = "linear")

mdl1 %>% compile(loss = 'mse', optimizer = 'rmsprop')

模型如下

___________________________________________________________
Layer (type)               Output Shape         Param #    
===========================================================
lstm_25 (LSTM)             (None, 1, 512)       1056768    
___________________________________________________________
dropout_25 (Dropout)       (None, 1, 512)       0          
___________________________________________________________
lstm_26 (LSTM)             (None, 512)          2099200    
___________________________________________________________
dropout_26 (Dropout)       (None, 512)          0          
___________________________________________________________
dense_13 (Dense)           (None, 1)            513        
===========================================================
Total params: 3,156,481
Trainable params: 3,156,481
Non-trainable params: 0
___________________________________________________________

为了训练模型，提前停止与验证集一起使用。

mdl1.history <- mdl1 %>% 
  fit(dt.tr, dt.tr.out, epochs=500, shuffle=F,
      validation_data = list(dt.val, dt.val.out),
      callbacks = list(
        callback_early_stopping(min_delta = 0.000001,  patience = 10, verbose = 1)
      ))

最重要的是，我想使用批量归一化来加速训练。根据我的理解，要使用批量归一化，我需要将数据分成批次，并应用于layer_batch_normalization每个隐藏层的输入。模型层如下所示：

batch_size <- 32
mdl2 <- keras_model_sequential()
mdl2 %>%
  layer_batch_normalization(input_shape = c(1, 3), batch_size = batch_size) %>%

  layer_lstm(units = 512, return_sequences = T) %>%
  layer_dropout(rate = 0.3) %>%
  layer_batch_normalization(batch_size = batch_size) %>%

  layer_lstm(units = 512, return_sequences = F) %>%
  layer_dropout(rate = 0.2) %>%
  layer_batch_normalization(batch_size = batch_size) %>%

  layer_dense(units = 1, activation = "linear")

mdl2 %>% compile(loss = 'mse', optimizer = 'rmsprop')

该模型如下所示：

______________________________________________________________________________
Layer (type)                                    Output Shape       Param #    
==============================================================================
batch_normalization_34 (BatchNormalization)     (32, 1, 3)         12         
______________________________________________________________________________
lstm_27 (LSTM)                                  (32, 1, 512)       1056768    
______________________________________________________________________________
dropout_27 (Dropout)                            (32, 1, 512)       0          
______________________________________________________________________________
batch_normalization_35 (BatchNormalization)     (32, 1, 512)       2048       
______________________________________________________________________________
lstm_28 (LSTM)                                  (32, 1, 512)       2099200    
______________________________________________________________________________
dropout_28 (Dropout)                            (32, 1, 512)       0          
______________________________________________________________________________
batch_normalization_36 (BatchNormalization)     (32, 1, 512)       2048       
______________________________________________________________________________
dense_14 (Dense)                                (32, 1, 1)         513        
==============================================================================
Total params: 3,160,589
Trainable params: 3,158,535
Non-trainable params: 2,054
______________________________________________________________________________

训练模型看起来像以前一样。唯一的区别在于训练和验证数据集，它们的大小是 的倍数batch_size（此处为 32），通过从最后第二批到最后一批的数据重新采样来组成。

然而， 的性能mdl1比 的好得多mdl2，如下所示。

我不确定我到底做错了什么，因为我是从 keras（以及一般的实用神经网络）开始的。另外，第一个模型的性能也不是很好；任何有关如何改进的建议也很好。

Answer 1

我将 Keras 与 Python 一起使用，但我可以尝试 R。在该fit方法中，文档说如果省略，则默认为 32。在当前版本中，情况不再如此，因为可以在源代码中看到。我认为你应该这样尝试，至少在Python中是这样的：

mdl2 <- keras_model_sequential()
mdl2 %>%
  layer_input(input_shape = c(1, 3))  %>%

  layer_batch_normalization() %>%
  layer_lstm(units = 512, return_sequences = T, dropout=0.3) %>%

  layer_batch_normalization() %>%
  layer_lstm(units = 512, return_sequences = F, dropout=0.2) %>%

  layer_batch_normalization() %>%
  layer_dense(units = 1, activation = "linear")

mdl2 %>% compile(loss = 'mse', optimizer = 'rmsprop')
mdl2.history <- mdl2 %>% 
  fit(dt.tr, dt.tr.out, epochs=500, shuffle=F,
      validation_data = list(dt.val, dt.val.out),
      batch_size=32,
      callbacks = list(
        callback_early_stopping(min_delta = 0.000001,  patience = 10, verbose = 1)
      ))