Prediction of Compressive Strength of Concrete and Rock Using an Elementary Instance-Based Learning Algorithm

Table 6

Prediction performances of individual models.

Dataset	ML technique	RMSE	R^2	Literature
1	Concrete compressive strength
	ANN	5.867	0.993	Chou & Pham [11]
	CART	7.860	0.983
	CHAID	8.307	0.980
	MLR	11.735	0.961
	GENLIN	11.596	0.962
	SVM	13.145	0.950
	4-NN (RepeatedCV)	10.693	0.808
	4-NN (CV)	10.782	0.805
	5-NN (LOOCV)	10.829	0.803
2	Concrete compressive strength
	ANN	7.104	0.832	Chou & Pham [11]
	CART	7.312	0.848
	CHAID	10.040	0.699
	MLR	5.161	0.917
	GENLIN	5.163	0.917
	SVM	10.954	0.726
	7-NN (CV)	7.437	0.756
	8-NN (RepeatedCV)	7.525	0.762
	8-NN (LOOCV)	7.734	0.743
3	Concrete compressive strength
	ANN	1.548	0.987	Chou & Pham [11]
	CART	1.837	0.982
	CHAID	1.994	0.979
	MLR	1.996	0.979
	GENLIN	1.996	0.980
	SVM	2.012	0.981
	Stepwise regression	2.020	0.956	Ahmadi-Nedushan [38]
	2-NN (LOOCV)	1.747	0.966
	2-NN (CV)	1.834	0.967
	2-NN (RepeatedCV)	1.856	0.965
4	Rock compressive strength
	ANN	N/A	0.921	Teymen and Mengüç [27]
	MRA	N/A	0.953
	8-NN (CV)	21.954	0.896
	9-NN (RepeatedCV)	21.956	0.899
	9-NN (LOOCV)	22.476	0.887
5	Granite compressive strength
	ANN	26.203	0.804	Jahed Armaghani et al. [13]
	MRA	13.818	0.891
	3-NN (CV)	14.509	0.885
	3-NN (RepeatedCV)	14.867	0.872
	3-NN (LOOCV)	15.576	0.862
	Granite Young’s modulus
	ANN	21.022	0.643	Jahed Armaghani et al. [13]
	MRA	22.192	0.596
	7-NN (CV)	21.850	0.630
	7-NN (RepeatedCV)	21.876	0.638
	7-NN (LOOCV)	22.654	0.580
6	Rock compressive strength
	ANN	7.58	0.96	Jalali et al. [30]
	MRA	10.80	0.91	Heidari et al. [33]
	2-NN (LOOCV)	10.387	0.926
	2-NN (RepeatedCV)	10.466	0.929
	2-NN (CV)	10.758	0.927

Highlighted ones in bold denote the k-NN model and performance measure.