Table 2 Performance of artificial neural network models with various numbers of layers.
From: Performance and efficiency of machine learning algorithms for analyzing rectangular biomedical data
Hidden layer settings | Neural network algorithm | Total accuracy (%) | MSE |
|---|---|---|---|
Single hidden layer | |||
5 | Trainscg | 69.61 | 0.0852 |
10 | Trainscg | 69.34 | 0.0857 |
20 | Trainscg | 69.88 | 0.0846 |
40 | Trainscg | 70.10 | 0.0852 |
50 | Trainscg | 70.12 | 0.0852 |
60 | Trainscg | 69.45 | 0.0861 |
70 | Trainscg | 69.54 | 0.0860 |
80 | Trainscg | 69.37 | 0.0867 |
90 | Trainscg | 68.88 | 0.0876 |
100 | Trainscg | 69.10 | 0.0865 |
Double hidden layers | |||
[5,5] | Trainscg | 69.63 | 0.0857 |
[10] | Trainscg | 69.72 | 0.0853 |
[20] | Trainscg | 69.83 | 0.0851 |
[40] | Trainscg | 70.16 | 0.0852 |
[50] | Trainscg | 69.88 | 0.0854 |
[60] | Trainscg | 65.06 | 0.1398 |
[70] | Trainscg | 69.70 | 0.0863 |
[80] | Trainscg | 69.48 | 0.0885 |
[90] | Trainscg | 69.21 | 0.0882 |
[100,100] | Trainscg | 69.28 | 0.0867 |
