Skip to main content

Table 1 Network parameter settings of the SAE

From: Automatic classification of nerve discharge rhythms based on sparse auto-encoder and time series feature

Parameter name

Parameter value

Number of input layer nodes

1024

Number of output layer nodes

4

Number of nodes in layer 1

1224

Number of nodes in layer 2

824

Unsupervised training epochs

1000

Supervised training epochs

1000

L2 weight regularization

0.01

Loss function

Crossentropy

Sparsity regularization β

0.1

Training algorithm

Trainscg