Skip to main content

Advertisement

Table 6 False negative prediction: statistical significance of the difference in performance between the methods.

From: A comparison of machine learning techniques for survival prediction in breast cancer

ANOVA
P= 2.75 × 10-9
GP vs. SVM-K1 GP vs. SVM-K2 GP vs. SVM-K3 GP vs. MP GP vs. RF
P = 2.74 × 10-6 P = 3.32 × 10-6 P = 1.27 × 10-10 P = 8.53 × 10-6 P = 4.65 × 10-7
  1. First line shows ANOVA test on the 6 samples of solutions found by each method, while second line depicts pairwise 2-tailed Student t-tests comparing GP with each other method.