Conference Proceeding
Details
Citation
Brownlee A, Wallace A & Cairns D (2021) Mining Markov Network Surrogates to Explain the Results of Metaheuristic Optimisation. In: Martin K, Wiratunga N & Wijekoon A (eds.) Proceedings of the SICSA eXplainable Artifical Intelligence Workshop 2021. CEUR Workshop Proceedings, 2894. SICSA eXplainable Artifical Intelligence Workshop 2021, Aberdeen, 01.06.2021-01.06.2021. Aachen: CEUR Workshop Proceedings, pp. 64-70. http://ceur-ws.org/Vol-2894/short9.pdf
Abstract
Metaheuristics are randomised search algorithms that are
effective at finding ”good enough” solutions to optimisation problems.
However, they present no justification for the generated solutions, and are non-trivial to analyse. We propose that identifying which combinations of variables strongly influence solution quality, and the nature of that relationship, represents a step towards explaining the choices made by the algorithm. Here, we present an approach to mining this information from a “surrogate fitness function” within a metaheuristic. The approach is demonstrated with two simple examples and a real-world case study.
Keywords
metaheuristics; surrogates; optimisation; explainability;
Status | Published |
---|---|
Title of series | CEUR Workshop Proceedings |
Number in series | 2894 |
Publication date | 31/12/2021 |
Publication date online | 30/06/2021 |
URL | http://hdl.handle.net/1893/33484 |
Publisher | CEUR Workshop Proceedings |
Publisher URL | http://ceur-ws.org/Vol-2894/short9.pdf |
Place of publication | Aachen |
ISSN of series | 1613-0073 |
Conference | SICSA eXplainable Artifical Intelligence Workshop 2021 |
Conference location | Aberdeen |
Dates |
People (3)
Senior Lecturer in Computing Science, Computing Science and Mathematics - Division
Lecturer, Computing Science
Tutor, Computing Science