Table 2. Performance of the predictions under the combinations of RF with six feature scoring methods.

Machine learning method	Feature scoring method	Accuracy (%)	Sensitivity (%)	Specificity (%)	Precision (%)	MCC	F1	AUCROC
Random forest	Fscore	74.79	76.27	73.31	74.07	0.4960	0.7516	0.8202
	Kmeans	72.88	73.31	72.46	72.69	0.4576	0.7300	0.7933
	Lasso	73.09	76.69	69.49	71.54	0.4631	0.7403	0.8087
	Pearson	77.12	80.51	73.73	75.40	0.5436	0.7787	0.8193
	Spearman	76.27	79.66	72.88	74.60	0.5266	0.7705	0.8163
	Ttest	74.15	78.39	69.92	72.27	0.4848	0.7520	0.8046

Note:

The highest metric is highlighted in bold.