Rainfall modeling using two different neural networks improved by metaheuristic algorithms

Environmental Sciences Europe

Table 4 Statistical indexes for different soft computing models

Index	MAE, mm	NSE	PBIAS
Station (1) (Train)
MLP–HGSO	0.687	0.95	0.17
MLP–PSO	0.727	0.89	0.22
MLP–BA	0.825	0.91	0.19
RBFNN–HGSO	0.722	0.93	0.16
RBFNN–BA	0.924	0.87	0.26
RBFNN–PSO	0.941	0.84	0.28
Station (1) (Test)
MLP–HGSO	0.712	0.90	0.23
MLP–PSO	0.755	0.83	0.29
MLP–BA	0.765	0.85	0.25
RBFNN–HGSO	0.717	0.87	0.27
RBFNN–BA	0.865	0.81	0.31
RBFNN–PSO	0.891	0.80	0.35
Station (2) (Train)
MLP–HGSO	0.683	0.97	0.15
MLP–PSO	0.725	0.90	0.21
MLP–BA	0.723	0.92	0.20
RBFNN–HGSO	0.691	0.93	0.17
RBFNN–BA	0.914	0.89	0.22
RBFNN–PSO	0.930	0.86	0.23
Station (2) (Test)
MLP–HGSO	0.711	0.92	0.22
MLP–PSO	0.743	0.85	0.28
MLP–BA	0.742	0.89	0.26
RBFNN–HGSO	0.719	0.91	0.25
RBFNN–BA	0.863	0.83	0.31
RBFNN–PSO	0.890	0.82	0.32