Table 3.

Average RMSE, MAE, and $R^2$ of six algorithms for yield prediction. A 10-fold cross-validation on the training dataset was used for algorithm performance evaluation, since the ground truth yield of the test dataset was never released.

Model	Train			Validation
	RMSE	MAE	$R^2$	RMSE	MAE	$R^2$
Linear regression	0.1016	0.1009	0.1047	0.1026	0.0851	0.0866
Factorization machine	0.0740	0.0676	0.4855	0.0984	0.0765	0.1578
Xgboost	0.0790	0.0735	0.4581	0.0996	0.0806	0.1388
G $\times$ E	0.0740	0.0706	0.4902	0.0980	0.0744	0.1623
Random forest	0.0737	0.0673	0.5283	0.0976	0.0723	0.1738
Proposed model	0.0548	0.0523	0.7386	0.0869	0.0648	0.3448