Skip to main content
. 2022 Feb 5;195:116611. doi: 10.1016/j.eswa.2022.116611

Table 2.

Hyper parameter and structure of CNN, and CNN+LSTM models.

Models Layers Number kernel Bias Kernel
of filters/ size regularizer regularizer
units L1 L1
CNN Conv1D 100 2 0.4 0.002
Conv1D 75 2 0.4 0.002
MaxPooling1D 2
Flatten
Dense 64
Dense 1
CNN+LSTM Conv1D 100 2 0.002
MaxPooling1D 2
Flatten
LSTM 64 0.5
Dropout 0.3
Dense 1