We finished the part dedicated to Empirical Analysis by describing multiple comparisons and sequential testing procedures such as the racing algorithm. We also discussed issues related with the design of experiments on algorithms.
The example file for the race as well as few lines for producing plots in R are available from the section Resources. Other functions demonstrated at the lecture were expand.grid, pairwise.t.test, pairwise.wilcox.test, plot(TukeyHSD(aov(...))).
In the last lecture we will go through an example study and mention hybrid metaheuristics, multiobjective optimization and optimization under uncertainty. The article to follow the example study has been distributed at the lecture, the rest will be on the slides.
We will conclude evaluating the course. Prepare
constructive critics!