Tutorial on Benchmarking Multiobjective Optimizers 2.0

D. Brockhoff, T. Tušar

In this tutorial, we discuss the past and future of benchmarking multiobjective optimizers. In particular, we discuss the new view on benchmarking multiobjective algorithms by falling back on single-objective comparisons and thus being able to use all methodologies and tools from the single-objective domain such as empirical distributions of runtimes. We also discuss the advantages and drawbacks of some widely used multiobjective test suites and explain how we can do better, going back to the roots of what a multi-objective problem is in practice, namely the simultaneous optimization of multiple objective functions. We also discuss recent advances in the visualization of (multiobjective) problem landscapes and compare the previous and newly proposed benchmark problems in the context of those landscape visualizations.

The tutorial was held at GECCO 2021 and GECCO 2022.

Download tutorial slides from here.