Optimising the Selection of Input Variables to Increase the Predicting Accuracy of Shear Strength for Deep Beams

Table 4

The statistical description of the proposed models: first scenario.

Model	Training phase				Testing phase
	MAE (kN)	RMSE (kN)	NSE	WI	MAE	RMSE (kN)	NSE	WI
LWLR	13.249	22.563	0.974	0.993	33.933	57.776	0.941	0.984
RF	18.138	28.613	0.958	0.988	49.068	77.712	0.892	0.964
MLR	36.924	48.692	0.879	0.967	61.165	89.651	0.857	0.953
ELM	23.194	31.045	0.951	0.987	51.999	78.200	0.891	0.968