Table 4.

Sensitivity analysis of the error with respect to the hyperparameter $\beta$ and the number of iterations used in the training of the view models considering the database CIFAR-10. The best accuracy value is in bold.

$\mathit{\beta }$/Epoch	20	40	60	100
0.001	20.23	22.11	21.82	29.03
0.01	12.94	11.32	11.38	18.23
0.02	12.02	10.44	11.36	17.17
0.03	12.78	11.39	11.38	15.95
0.04	13.15	11.44	11.40	12.33
0.05	16.36	16.10	15.87	15.88
0.06	19.13	18.01	18.29	20.44
0.07	19.15	19.23	19.55	23.04
0.08	20.59	21.65	21.90	25.70
0.09	23.44	22.63	22.06	25.33
0.11	24.73	23.24	22.09	25.07
0.12	24.81	23.31	22.16	25.20
0.13	24.91	23.47	22.50	25.53
0.14	25.25	23.93	22.84	25.72
0.15	25.53	24.02	23.29	27.12
0.16	25.90	24.30	23.42	28.70
0.2	35.66	34.58	35.87	35.99
0.3	58.67	54.85	57.90	52.01
0.5	68.33	68.95	68.41	68.11
0.6	67.07	66.91	66.02	65.61
0.7	71.11	70.92	70.77	70.89