Table 4 Random data tests. The experimentally determined largest number of pairs found from 20 runs on random data. Each class has half the number of samples (number positives = number of negatives). The total number of genes is 2000. The second column is the number of pairs found for this 2000 gene set size, third column is the observed probability a single pair will be a perfect classifier ("observed" / 2000²/2).

From: Many accurate small-discriminatory feature subsets exist in microarray transcript data: biomarker discovery