Developing a Robust Model Based on the Gaussian Process Regression Approach to Predict Biodiesel Properties

Table 4

The comparison between four kernel functions for PP property from the point of view of statistical quantities.

Model	Phase	R^2	MRE (%)	MSE	RMSE	STD
GPR (Exponential)	Train	0.982	0.00056	0.00000	0.00195	0.00124
	Test	0.878	0.73386	6.64170	2.57715	1.73628
	Total	0.967	0.17651	1.59401	2.57715	1.18811
GPR (Matern)	Train	0.989	0.21290	0.54286	0.73679	0.46621
	Test	0.889	0.55839	2.85624	1.69004	0.80875
	Total	0.973	0.30712	1.15081	1.69004	0.68377
GPR (Squared Exponential)	Train	1.000	0.26262	0.86125	0.92803	0.60432
	Test	0.968	0.52306	2.43573	1.56068	0.72847
	Total	0.966	0.33494	1.33190	1.56068	0.72451
GPR (Rational Quadratic)	Train	0.984	0.25484	0.73098	0.85497	0.51735
	Test	0.788	0.73120	6.62593	2.57409	1.78533
	Total	0.954	0.31634	1.90128	2.57409	1.09967