Table 10. Training Functions for Hyperparameter Groups of OPR, WOR, and GOR.

training function (backpropagation)	production rate	WOR	GOR
Levenberg–Marquardt (trainlm)	1, 16, 31, 23–30,	1, 16, 31, 23–30,	1, 8–16, 23–31, 38–45
Bayesian regularization (trainbr)	2, 17, 32, 8–15, 38–45	2, 17, 32,	2, 17, 32,
resilient backpropagation (trainrp)	3, 18, 33	3, 18, 33, 8–15, 38–45	3, 18, 33
BFGS quasi-Newton (trainbfg)	4, 19, 34	4, 19, 34	4, 19, 34
conjugate gradient with Fletcher–Reeves updates (traincgf)	5, 20, 35	5, 20, 35	5, 20, 35
conjugate gradient with Powell–Beale restarts (traincgb)	6, 21, 36	6, 21, 36	6, 21, 36
conjugate gradient with Polak–Ribiére updates (traincgp)	7, 22, 37	7, 22, 37	7, 22, 37