Table 7 Comparison of CNNs with the different number of cells.

From: Light convolutional neural network by neural architecture search and model pruning for bearing fault diagnosis and remaining useful life prediction

Model

Parameter size (MB)

Training accuracy (%)

Validation accuracy (%)

Test accuracy (%)

2 cells

0.197

99.672

99.954

99.969

3 cells

0.232

99.583

99.956

100.000

4 cells

0.589

99.469

99.957

99.968

5 cells

0.706

99.469

99.956

100.000

6 cells

0.737

98.491

100.000

99.937

Base CNN

9.677

98.491

100.000

99.937