Wilcoxon Rank Sum Test

I have minimum, average and maximum values of an minimisation objective obtained by two different algorithms. Now to test the robustness of those algorithms how to run Wilcoxon Rank Sum Test?

[David Hill]

Did you look at the documentation for the matlab function ranksum ?

[Adam Danz]

What David Hill said...

Here's a list of non-parametric ttests

For an independent sample ttest **with equal variances**,

• Mann-Whitney U-Test (in matlab: ranksum(x,y))

For an independent sample ttest **with unequal variances**

• Kolmogorov-smirnov test (in matlab: kstest2(x,y) for 2-sample and ktest(x) for 1 sample

For a paired ttest

• Wilcoxon Signed-Rank test (in matlab: ranksum(x,y))