Table 2.

Overview of the literature using the UCI-concrete dataset. SVM (Support Vector Machine), GBT (Gradient Boosted Trees), BNN (Bagged NN), GBNN (Gradient Boosted NN), WBNN (Wavelet Bagged NN), WGBNN (Wavelet Gradient Boosted NN), Regression Forest (RF), Decision tree (DT), Gridsearch (GS), Gridsearch + Feature construction (GS+).

Source	Year	Model	Param. Tuning	Train-Test-Split	$R^2$	RMSE
Yeh30	1998	Linear Regression	Manual	75/25	0.770	N/A
Yeh	1998	Neural Network (NN)	Manual	75/25	0.914	N/A
Chou45	2011	SVM	Manual	10-fold CV − avg	0.8858	5.619
Chou	2011	GBT	Manual	10-fold CV − avg	0.9108	4.949
Erdal46	2013	BNN	GS	90/10	0.9278	4.870
Erdal	2013	GBNN	GS	90/10	0.9270	5.240
Erdal	2013	WGBNN	GS	90/10	0.9528	5.750
Golafshani47	2019	Symbolic Regression	Manual	75/25	0.8008	10.6984
Han43	2019	RF + 1 const. Feat.	GS+	$50\times 90$/10 split − avg	0.9322	4.434
Nguyen-Sy48	2020	GBT [XGBoost]	GS	10-fold CV − avg	0.93	4.270
Feng42	2020	AdaBoost with DT	GS	10-fold CV − avg	0.952	4.856
Chakraborty44	2021	GBT [XGBoost]	GS + RFE	90/10	0.979	2.650