Hyperparameter Tuning of Machine Learning Algorithms Using Response Surface Methodology: A Case Study of ANN, SVM, and DBN

Table 13

Analysis of variance of the second fold.

Source	Df	Adj SS	Adj MS	F-value	value
Model	8	363.18	45.397	14.5	<0.001
HN	1	0.446	0.446	0.14	0.708
LRRBM	1	0.652	0.652	0.21	0.650
LR	1	137.73	137.73	43.98	<0.001
N	1	14.968	14.968	4.78	0.034
ni	1	153.99	153.99	49.17	<0.001
Nb	1	28.911	28.911	9.23	0.004
HN × HN	1	11.969	11.969	3.82	0.057
LR × n	1	14.51	14.51	4.63	0.037
Error	44	137.8	3.132
Lack-of-fit	36	111.06	3.085	0.92	0.603
Pure error	8	26.738	3.342
Total	52	500.98