Table 3.
The best 10 results (test #1, ear).
| Trial | Images | Number of hidden layers and number of neurons | Persons per module | Rec. rate |
Error | |
|---|---|---|---|---|---|---|
| Training | Testing | |||||
| 1 | 80% (1, 2, and 3) |
20% (4) |
5 (126, 96, 179, 239, 37) 4 (188, 196, 93, 171) 5 (109, 107, 110, 168, 29) |
Module #1 (1 to 12) Module #2 (13 to 40) Module #3 (41 to 77) |
100% (77/77) |
0 |
|
| ||||||
| 2 | 69% (2, 3 and 4) |
31% (1) |
5 (222, 238, 113, 27, 75) 4 (151, 53, 99, 79) 2 (209, 31) 2 (144, 71) 4 (30, 218, 194, 199) 4 (25, 81, 239, 20) 5 (237, 43, 83, 102, 128) |
Module #1 (1 to 5) Module #2 (6 to 21) Module #3 (22 to 31) Module #4 (32 to 46) Module #5 (47 to 63) Module #6 (64 to 73) Module #7 (74 to 77) |
100% (77/77) |
0 |
|
| ||||||
| 3 | 66% (2, 3, and 4) |
34% (1) |
5 (141, 70, 120, 158, 242) 4 (124, 55, 23, 243) 3 (96, 186, 213) 4 (28, 62, 51, 42) 1 (223) |
Module #1 (1 to 34) Module #2 (35 to 40) Module #3 (41 to 44) Module #4 (45 to 75) Module #5 (76 to 77) |
100% (77/77) |
0 |
|
| ||||||
| 4 | 74% (2, 3, and 4) |
26% (1) |
5 (139, 97, 200, 121, 231) 5 (204, 114, 164, 216, 138) 5 (195, 137, 124, 71, 86) 5 (144, 70, 92, 220, 63) 5 (119, 176, 154, 167, 161) 4 (199, 162, 96, 65) |
Module #1 (1 to 6) Module #2 (7 to 29) Module #3 (30 to 50) Module #4 (51 to 58) Module #5 (59 to 71) Module #6 (72 to 77) |
100% (77/77) |
0 |
|
| ||||||
| 5 | 63% (2, 3, and 4) |
37% (1) |
5 (136, 183, 149, 193, 161) 5 (181, 132, 175, 140, 155) |
Module #1 (1 to 68) Module #2 (69 to 77) |
100% (77/77) |
0 |