Table 3.

The performance of regression learning algorithms for conductivity estimation.

Algorithm name	Model type	RMSE	R2
Linear regression	Linear	1.29	0.43
Linear regression	Interactions linear	0.99	0.66
Linear regression	Robust linear	1.65	0.07
Linear regression	Stepwise LINEAR	0.99	0.66
Tree	Fine tree	0.37	0.95
Tree	Medium tree	0.58	0.88
Tree	Coarse tree	0.81	0.77
SVM	Linear SVM	1.43	0.30
SVM	Quadratic SVM	0.95	0.69
SVM	Cubic SVM	5.48	0.00
SVM	Fine Gaussian SVM	0.48	0.92
SVM	Medium Gaussian SVM	0.67	0.84
SVM	Coarse Gaussian SVM	1.39	0.34
Ensemble	Boosted trees	0.41	0.94
Ensemble	Bagged trees	0.53	0.90
GPR	Squared exponential GPR	0.27	0.97
GPR	Matern 5/2 GPR	0.25	0.98
GPR	Rational quadratic GPR	0.25	0.98
GPR	Exponential GPR	0.15	0.99