Comparison of performance of PMT-UC, PMG-UC and PMT-DC applied on binary-clusters simulated datasets.

The clustering and gene selection results for the four set-ups with , in terms of the average frequencies (N) of the selected numbers of clusters (K), and the average of RI, aRI, FN, and TN in 50 simulations, where FN is the number of informative variables incorrectly selected to be noninformative and TN is the number of noninformative variables correctly selected. In the true case, , , . The table indicates in bold all results that perform best or that are not significantly different from each other.‡: denotes that can be 4 or 5.†: denotes that can be any element of the set which contains the predefined numbers of clusters.