On the Training Algorithms for Artificial Neural Network in Predicting the Shear Strength of Deep Beams
Table 7
Final values of weights and biases of the typical ANN-CG model (Case 2).
IW (10 × 9)
LW (1 × 10)
BL (1 × 10)
BO (1 × 1)
−0.576
−0.075
−0.679
0.078
−0.113
−0.950
1.181
−0.257
0.343
0.410
1.589
−0.390
0.367
0.052
0.518
−0.424
−0.739
0.350
0.784
0.819
−0.877
−0.285
−1.485
—
1.281
−0.106
−0.286
0.648
−0.339
0.771
−0.897
0.585
−0.540
0.645
−0.663
—
0.278
0.060
−1.788
0.298
−0.343
−0.402
0.014
0.061
0.887
0.976
−0.924
—
−0.469
−0.099
0.137
−0.538
1.265
−0.285
0.417
−0.103
−1.010
0.307
0.384
—
−0.036
−0.891
0.371
−0.054
0.743
0.517
−0.866
−0.274
0.671
−0.083
−0.067
—
−0.440
0.312
−1.232
−0.598
−0.621
−1.295
0.010
−0.475
−0.229
0.019
−0.601
—
−0.290
−0.348
−0.482
0.505
1.006
−0.248
−0.507
−0.133
0.930
0.234
−1.274
—
0.395
0.257
−0.233
−0.362
0.718
−0.166
0.734
−0.552
−1.248
0.471
1.490
—
−0.311
0.658
0.839
−0.248
0.276
0.947
0.875
0.519
−0.070
0.398
−1.798
—
Note. IW = the input to hidden layer weights; LW = the hidden layer to output layer weights; BL = the input to hidden layer biases; BO = the hidden layer to output layer bias.