Figure 2
From: Reordering based integrative expression profiling for microarray classification
Microarray classification performance comparisons for different integrative expression profiling (IXP) approaches. Using GSE1297 as training set (31 samples, 22 AD patients vs. 9 controls), and GSE5281 (151 samples, 84 AD patients vs. 67 controls) as testing set, two-class (AD vs. control) classifications are performed for IXP approaches, based on ant colony optimization reordering (ACOR) algorithm, network ranking, graph clustering, and random-ordering (RandRank), with different coefficient r in Equation (2). UniWeight: unified gene weights (all equal to one).