In this chapter, a scientific tool designed to facilitate fair comparisons of heuristics is introduced. Making a fair comparison of the performance of different algorithms is a general problem for the heuristic community. Most of the works on experimental analysis of heuristic algorithms have been focused on tabular comparisons of experimental results over standard sets of benchmark instances. However, from a statistical point of view, and according to the experimental design theory, a minimum requirement to compare heuristic algorithms is the use of non-parametric tests. Non-parametric tests can be used for comparing algorithms whose results represent average values, in spite of the inexistence of relationships between them, and explicit conditions of normality, among others. The proposed tool, referred to as VisTHAA, incorporates four non-parametric statistical tests to facilitate the comparative analysis of heuristics. As a case study, VisTHAA is applied to analyze the published results for the best state-of-the-art algorithms that solve the one-dimensional Bin Packing Problem.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.