Table 1.

Comparison of model performance for the prediction of BCF in Cyprinus carpio. MAE is the mean absolute error and NA indicates the metric was not applicable.

	Model	RMSE			R2			MAE
		Training	Verification	Test	Training	Verification	Test	Training	Verification	Test
Trajan	Linear	0.785	1.052	0.832	0.532	0.390	0.521	0.619	0.835	0.608
	GRNN	0.830	0.893	0.873	0.673	0.400	0.569	0.664	0.893	0.718
	RBF	0.723	0.689	0.584	0.651	0.635	0.725	0.565	1.600	0.450
	3-MLP	0.689	0.538	0.337	0.675	0.770	0.659	0.548	1.608	0.553
	4-MLP	0.403	0.524	0.644	0.887	0.819	0.702	0.313	0.380	0.530
	Model	Training	Cross-Validation	Test	Training	Cross-Validation	Test	Training	Cross-Validation	Test
R	OLM	0.719	0.771	1.203	0.621	0.570	0.234	0.560	NA	0.778
	PLS	0.722	0.769	1.164	0.618	0.571	0.254	0.564	NA	0.765
	RR	0.725	0.766	1.083	0.614	0.576	0.304	0.568	NA	0.753
	EN	0.729	0.760	1.054	0.612	0.582	0.314	0.577	NA	0.754
	QRL	0.733	0.757	1.112	0.607	0.585	0.284	0.562	NA	0.770
	KNN	0.517	0.683	0.902	0.807	0.665	0.468	0.404	NA	0.648
	ELM	0.673	0.756	1.014	0.668	0.593	0.346	0.529	NA	0.768
	ANN-1HL	0.596	0.751	0.877	0.739	0.597	0.505	0.462	NA	0.620
	ANN-a1HL	0.395	0.672	0.859	0.888	0.678	0.518	0.319	NA	0.612
	ANN-2HL	0.232	0.834	1.022	0.962	0.560	0.370	0.174	NA	0.680
	ANN-3HL	0.454	0.795	0.880	0.860	0.582	0.520	0.345	NA	0.624
	MARS	0.539	0.730	1.014	0.787	0.632	0.390	0.425	NA	0.696
	B-MARS	0.500	0.681	0.899	0.819	0.673	0.479	0.395	NA	0.633
	SVM-R	0.383	0.644	0.841	0.893	0.704	0.537	0.261	NA	0.590
	SVM-P	0.699	0.747	1.029	0.643	0.594	0.340	0.539	NA	0.729
	RF-CART	0.292	0.675	0.771	0.956	0.688	0.633	0.231	NA	0.589
	RF-CIT	0.605	0.739	0.821	0.762	0.630	0.586	0.485	NA	0.652
	BT	0.249	0.660	0.789	0.957	0.687	0.593	0.187	NA	0.587
	CR	0.353	0.678	0.973	0.910	0.673	0.431	0.282	NA	0.628