Forward and Reverse Process Models for the Squeeze Casting Process Using Neural Network Based Approaches

Table 2

BPNN parametric study results of both forward and reverse mappings.

NN parameters	Forward mapping	Error	Reverse mapping	Error
Hidden neurons	11	0.017321	18	0.034095
Learning rate-hidden layer,	0.01	0.008271	0.5885	0.033450
Learning rate-output layer,	0.5885	0.003099	0.2325	0.033410
Momentum constant,	0.455	0.003099	0.455	0.033410
Activation constant-hidden layer	2.8	0.002525	5.5	0.033410
Activation constant 1-output layer	5.5	0.002525	5.5	0.033410
Activation constant 2-output layer	8.65	0.001984	5.5	0.033410
Bias	0.0000455	0.001458	0.0000505	0.033410