Table 2.

The comparison triplets between algorithm pairs from McTwo, CFS, PAM and RRF

CT(A, B)	McTwo	CFS	PAM	RRF
McTwo	0/17/0	1/2/14	12/1/4	15/0/2
CFS	14/2/1	0/17/0	16/1/0	17/0/0
PAM	4/1/12	0/1/16	0/17/0	13/0/4
RRF	2/0/15	0/0/17	4/0/13	0/17/0

The comparison triplet CT(A, B) is defined to be the numbers of the 17 datasets where algorithm A performs better, equally well and worse, compared with algorithm B. The measurement mAcc is used for comparison. The column and row of CFS are highlighted in bold