Geometric Learning in Black-Box Optimization: A GNN Framework for Algorithm Performance Prediction
Automated algorithm performance prediction in numerical blackbox optimization often relies on problem characterizations, such as exploratory landscape analysis features. These features are typically used as inputs to machine learning models and are represented in a tabular format. However, such approaches often overlook algorithm configurations, a key factor influencing performance. The relationships between algorithm operators, parameters, problem characteristics, and performance outcomes form a complex structure best represented as a graph.
Authors
- Carola Doerr (ORCID: https://orcid.org/0000-0002-4981-3227)
- Sašo Džeroski (ORCID: https://orcid.org/0000-0003-2363-712X)
- Ana Kostovska (ORCID: https://orcid.org/0000-0002-5983-7169)
- Panče Panov (ORCID: https://orcid.org/0000-0002-7685-9140)
- Tome Eftimov (ORCID: https://orcid.org/0000-0001-7330-1902)
Institutions
- Centre National de la Recherche Scientifique (FR)
- Jožef Stefan Institute (SI)
- Sorbonne Université (FR)
Publication Details
- Journal
- Proceedings of the Genetic and Evolutionary Computation Conference Companion
- Published
- 2025-07-14
- DOI
- https://doi.org/10.1145/3712255.3726696
- Citations
- 1
- Primary Topic
- Advanced Image and Video Retrieval Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.43
Funders
- HORIZON EUROPE Widening participation and spreading excellence