Suitability Evaluation of Crop Variety via Graph Neural Network

Table 2

The performance comparison of traditional machine learning methods and neural network under two data initialization situations. Important data are mar—d in bold.

	KNN	LR	SVM	NB	RF	DT	MLP	RBFNN	GAT	GCN
Accuracy	0.647	0.676	0.671	0.567	0.670	0.553	0.684	0.681	0.731	0.748
Precision score	0.628	0.651	0.644	0.562	0.662	0.563	0.670	0.657	0.743	0.687
Recall score	0.870	0.880	0.892	0.946	0.812	0.583	0.834	0.872	0.708	0.911
F1-score	0.729	0.748	0.748	0.705	0.730	0.545	0.749	0.750	0.725	0.783
AUC score	0.623	0.654	0.647	0.526	0.655	0.483	0.673	0.661	0.732	0.748