Table 4.

Results of ML analysis (eight methods) of BPW indices for discriminating between robust + no-MetS and possible sarcopenia (dynapenia + presarcopenia + sarcopenia).

Accuracy	SVM	MLP	GNB	DT	RF	LR	LDA	KNN
1	62.87	62.64	56.45	58.87	66.72	61.66	77.51	64.23
2	63.42	58.13	50.51	61.17	68.17	55.25	76.34	57.98
3	74.57	68.29	56.10	62.29	70.86	72.29	71.05	55.33
Average	66.95	63.02	54.35	60.77	68.58	63.07	*74.97	59.18

AUC	SVM	MLP	GNB	DT	RF	LR	LDA	KNN
1	0.52	0.56	0.58	0.53	0.61	0.54	0.76	0.61
2	0.56	0.54	0.51	0.59	0.65	0.49	0.75	0.57
3	0.69	0.67	0.62	0.62	0.66	0.67	0.72	0.59
Average	0.59	0.59	0.57	0.58	0.64	0.57	*0.74	0.59

The accuracy is in %. The accuracy and AUC values are listed for the threefold cross validation. Asterisks indicate the highest average value. Both accuracy and AUC were highest for LDA.