Structural Bias
We are excited to announce the Structural Bias Tutorial at GECCO 2025.
Benchmarking heuristic algorithms is vital for understanding under which conditions and on what kind of problems certain algorithms perform well. Most benchmarks are performance-based, to test algorithm performance under a wide set of conditions. Resourceand behaviour-based benchmarks test resource consumption and algorithm behaviour. In this Tutorial, we focus on behaviour benchmarking of algorithms and more specifically, we focus on Structural Bias (SB). SB is a form of bias inherent to the iterative heuristic optimisers in the search space that also affects the performance of the optimisation algorithm. Detecting whether, when, and what type of SB occurs in a heuristic optimisation algorithm can provide guidance on what needs to be improved in these algorithms, as well as help to identify conditions under which such bias would not occur. In the tutorial, we start by defining the problem of detecting and identifying different types of structural bias, including many visual examples. We then introduce state-of-the-art methods for bias detection. We follow up with SB results for several well-known and popular optimisation heuristics, give insights and show best practices to avoid SB in algorithm development. We conclude with a live demo of the Python-based BIAS toolkit which analyses a few well-known optimisation heuristics. Participants will be provided with links to live tools, necessary code and data.
Organizers:
Anna V. Kononova
