Research Article
Design Feed Forward Neural Network to Solve Singular Boundary Value Problems
Table 4
Weight and bias of the network for different training algorithm Example
1.
(a) |
| Weights and bias for trainlm | Net.IW{1, 1} | Net.LW{2, 1} | Net.B{1} |
| 0.2858 | 0.0759 | 0.1299 | 0.7572 | 0.0540 | 0.5688 | 0.7537 | 0.5308 | 0.4694 | 0.3804 | 0.7792 | 0.0119 | 0.5678 | 0.9340 | 0.3371 |
|
|
(b) |
| Weights and bias for trainbfg | Net.IW{1, 1} | Net.LW{2, 1} | Net.B{1} |
| 0.4068 | 0.8334 | 0.2601 | 0.1126 | 0.4036 | 0.0868 | 0.4438 | 0.3902 | 0.4294 | 0.3002 | 0.3604 | 0.2573 | 0.4014 | 0.1403 | 0.2976 |
|
|
(c) |
| Weights and bias for trainbr | Net.IW{1, 1} | Net.LW{2, 1} | Net.B{1} |
| 0.1696 | 0.8803 | 0.4075 | 0.2788 | 0.4711 | 0.8445 | 0.1982 | 0.4040 | 0.6153 | 0.1951 | 0.1792 | 0.3766 | 0.3268 | 0.9689 | 0.8772 |
|
|