Table 2.

Figures of merit for the different models (unsupervised and supervised) using GC × GC-HRTOFMS and SIFT-MS.

	Unsupervised random forest		Supervised random forest
	GC × GC-HRTOFMS	SIFT-MS	GC × GC-HRTOFMS	SIFT-MS
Accuracy (%)	53.66	65.22	75.61	76.09
Sensitivity	0.39	0.40	0.72	0.75
Specificity	0.65	0.85	0.78	0.77
Positive predictive value	0.47	0.67	0.72	0.71
Negative predictive value	0.58	0.65	0.78	0.80