Comparison of Machine Learning Techniques for the Prediction of Compressive Strength of Concrete

Table 3

Results of training and validation of the DT model, RF model, and NN model for compressive strength of concrete.

Model	Curing age	Results of training set		Results of validation set		Results of testing set
		R^2	RMSE	R^2	RMSE	R^2	RMSE
DT	28 days	0.7167	3.0717	0.4534	3.7188	0.4388	3.9073
	56 days	0.8604	2.0111	0.7519	2.4596	0.8008	2.2514
	91 days	0.8785	1.8454	0.8835	1.7957	0.8270	1.8294
RF	28 days	0.9426	1.6433	0.7943	1.3400	0.6870	2.6912
	56 days	0.9670	1.1306	0.9343	1.7282	0.9058	1.1449
	91 days	0.9780	0.8957	0.9828	1.6123	0.9213	1.0700
NN	28 days	0.9599	0.7251	0.9476	0.8213	0.9460	0.8106
	56 days	0.9769	0.7176	0.9872	0.8099	0.9500	0.9100
	91 days	0.9770	0.8023	0.9824	0.9641	0.9562	0.9106