Prediction of Compressive Strength of Concrete in Wet-Dry Environment by BP Artificial Neural Networks

Table 5

Prediction performance of different patterns of this model.

	R^2	MAE (MPa)	MAPE (%)	RMSE (MPa)
Training database	0.9684	2.47	3.52	2.51
Validation database	0.9270	4.02	5.13	4.13
Testing database	0.9478	3.87	4.97	4.04
All databases	0.9620	2.65	3.42	3.15