Table 7.

Variable selection for prostate cancer data (Gleason 5 vs. Gleason 7)

	GDFS	CFS	rpart	Predominant glycans (GDFS method)
Peaks 1–14	✗	✗	✗
Peak 15	✓	✓	✗	FA2BG2S1, A3G3
Peak 16	✗	✗	✗
Peak 17	✗	✗	✗
Peak 18	✗	✗	✗
Peak 19	✓	✓	✗	A3G3S2
Peak 20	✗	✗	✗
Peak 21	✓	✓	✗	A3G3S3
Peak 22	✗	✗	✗
Peak 23	✓	✗	✗	A4G4S4
Peak 24*	✓	✓	✗	A4F1G4S4

Features selected from the prostate cancer dataset (Gleason 5 vs. Gleason 7 cases) by the proposed GDFS method (GDFS), correlation-based feature selection (CFS), and recursive partitioning (rpart). Features that were selected in 90% more of the cross-validation models are marked with $✓$. Also listed are the predominant glycan structures corresponding to each selected peak. Detailed N-glycan composition of human serum was described in Royle et al. [9], and these peaks were also assigned in Saldova et al. [24]. *Peak 24 was the one most frequently selected by rpart, but less than 80% of the time.