Skip to main content

Table 8 Pair-wise Wilcoxon signed-ranks test for the 21 DHFR group results (W is the test statistic)

From: Multi-task learning with a natural metric for quantitative structure activity relationship learning

Setting

W

p-value

STL vs feature-based MTL medians: 0.821 and 0.808

108

0.8117

STL vs instance-based MTL medians: 0.821 and 0.668

222

3.147e−05

Feature-based MTL vs instance-based MTL medians: 0.808 and 0.668

220

5.245e−05