Skip to main content

Table 6 Friedman rank data for k = 12 methods and n = 10 cell differentiation datasets (Zagar et al. [70])

From: Exact p-values for pairwise comparison of Friedman rank sums, with application to comparing classifiers

Method GDS 2431 GDS 2666 GDS 2667 GDS 2668 GDS 2669 GDS 2671 GDS 2672 GDS 586 GDS 587 GDS 2688 Rank sum excluding GDS2688
MCE-euclid-FC 1 2 1 6 6 1 1 10 8 1 36
PCA-FC 5 1 6 1 1.5 12 8 5.5 1 3 41
PLS-AREA 6.5 8 4 3 4.5 5 6 7.5 3 6 47.5
PCA-AREA 4 6.5 3 2 7 11 7 7.5 2 2 50
MCE-euclid-AREA 3 3.5 2 5 9 3.5 5 11 9 4 51
PLS-FC 9 5 8 4 1.5 3.5 12 5.5 5.5 5 54
SVMRank-FC 11 9 5 8 8 6 3 1 5.5 7 56.5
SVMRank-AREA 9 11 9 7 3 10 2 2 4 8 57
PLS-FC-time 9 3.5 11 11 4.5 8 10 3 10 9 70
PLS-AREA-time 6.5 6.5 12 12 12 9 4 4 7 10 73
Pathrecon 2 12 7 10 11 2 9 9 11   73
PCA-Markers 12 10 10 9 10 7 11 12 12   93
Exact p-values for MCE-euclid-FC vs PLS-AREA-time
  d k n k 1 n 1 k 2 n 2 unadjusted Bonferroni-adjusted
1 × N comparison N × N comparison
Excluding GDS2688 37 12 9      0.016 0.174 1
Including GDS2688 46    12 9 10 1 0.003 0.038 0.230
  1. Note: Bold figures indicate p-values ≤ .05