Design Feed Forward Neural Network to Solve Singular Boundary Value Problems

Table 8

Weight and bias of the network for different training algorithm Example 2.

(a)

Weights and bias for trainlm
Net.IW{1, 1}	Net.LW{2, 1}	Net.B{1}
0.7094	0.1626	0.5853
0.7547	0.1190	0.2238
0.2760	0.4984	0.7513
0.6797	0.9597	0.2551
0.6551	0.3404	0.5060

(b)

Weights and bias for trainbfg
Net.IW{1, 1}	Net.LW{2, 1}	Net.B{1}
0.7094	0.1626	0.5853
0.7547	0.1190	0.2238
0.2760	0.4984	0.7513
0.6797	0.9597	0.2551
0.6551	0.3404	0.5060

(c)

Weights and bias for trainbr
Net.IW{1, 1}	Net.LW{2, 1}	Net.B{1}
0.9357	0. 7406	0.2122
0.4579	0.7437	0.0985
0.2405	0.1059	0.8236
0.7639	0.6816	0.1750
0.7593	0.4633	0.1636