Table 5.

The best 10 results (test #2, ear).

Trial	Images		Number of hidden layers and number of neurons	Persons per module	Rec. rate	Error
	Training	Testing
2	43% (2 and 3)	57% (1 and 4)	5 (115, 49, 187, 122, 194) 5 (182, 139, 50, 217, 54) 5 (132, 182, 56, 187, 159) 5 (167, 132, 121, 123, 219) 4 (116, 195, 54, 174) 5 (157, 108, 166, 95, 88) 5 (116, 119, 76, 121, 94) 5 (102, 58, 69, 111, 42)	Module #1 (1 to 9) Module #2 (10 to 22) Module #3 (23 to 33) Module #4 (34 to 36) Module #5 (37 to 51) Module #6 (52 to 63) Module #7 (64 to 75) Module #8 (76 to 77)	96.75% (149/154)	0.0325
4	48% (2 and 3)	52% (1 and 4)	4 (98, 136, 165, 141) 3 (176, 104, 215) 4 (142, 222, 65, 28) 5 (97, 139, 129, 99, 28) 4 (225, 83, 188, 34)	Module #1 (1 to 26) Module #2 (27 to 39) Module #3 (40 to 55) Module #4 (56 to 65) Module #5 (66 to 77)	96.75% (149/154)	0.0325
7	49% (2 and 3)	51% (1 and 4)	5 (201, 84, 169, 113, 131) 5 (199, 189, 62, 159, 151) 5 (104, 129, 88, 166, 66) 5 (123, 96, 52, 26, 67) 5 (125, 141, 86, 77, 105) 5 (121, 145, 87, 122, 31) 5 (36, 126, 146, 143, 145) 5 (126, 140, 88, 173, 206)	Module #1 (1 to 5) Module #2 (6 to 17) Module #3 (18 to 32) Module #4 (33 to 34) Module #5 (35 to 40) Module #6 (41 to 51) Module #7 (52 to 63) Module #8 (64 to 77)	96.75% (149/154)	0.0325
8	39% (2 and 3)	61% (1 and 4)	5 (125, 75, 69, 114, 140) 5 (138, 157, 101, 164, 98) 5 (76, 78, 86, 135, 70) 4 (74, 53, 57, 73) 5 (123, 55, 75, 125, 143) 5 (99, 118, 149, 224, 67) 5 (130, 184, 156, 180, 153)	Module #1 (1 to 11) Module #2 (12 to 14) Module #3 (15 to 27) Module #4 (28 to 33) Module #5 (34 to 43) Module #6 (44 to 57) Module #7 (58 to 77)	96.75% (149/154)	0.0325
14	40% (2 and 3)	60% (1 and 4)	5 (58, 26, 159, 123, 106) 5 (157, 156, 197, 22, 112) 4 (215, 78, 97, 220) 5 (120, 68, 219, 194, 58) 5 (142, 185, 141, 33, 187) 5 (108, 160, 61, 100, 54)	Module #1 (1 to 12) Module #2 (13 to 20) Module #3 (21 to 40) Module #4 (41 to 52) Module #5 (53 to 66) Module #6 (67 to 77)	96.75% (149/154)	0.0325