Impact of Training Instance Selection on Automated Algorithm Selection Models for Numerical Black-box Optimization

Genetic and Evolutionary Computation Conference (GECCO), pp. 1007–1016

Abstract

The recently proposed MA-BBOB function generator provides a way to create numerical black-box benchmark problems based on the well-established BBOB suite. Initial studies on this generator highlighted its ability to smoothly transition between the component functions, both from a low-level landscape feature perspective, as well as with regard to algorithm performance. This suggests that MA-BBOB-generated functions can be an ideal testbed for automated machine learning methods, such as automated algorithm selection (AAS).

Authors 4

1. Konstantin Dietrich

   Center for Scalable Data Analytics and Artificial Intelligence · Technische Universität Dresden

   Affiliation as printed

   ScaDS.AI, Dresden, Germany

   TU Dresden, Dresden, Germany

2. Diederick Vermetten Aachen

   Leiden University

   Affiliation as printed

   Leiden University, Leiden, Netherlands

   LIACS - Leiden Institute of Advanced Computer Science [Leiden] (Niels Bohrweg 1 2333 CA Leiden - Netherlands)

3. Carola Doerr

   Centre National de la Recherche Scientifique · Sorbonne Université

   Affiliation as printed

   CNRS, Paris, France

   Sorbonne Université, Paris, France

4. Pascal Kerschke

   Center for Scalable Data Analytics and Artificial Intelligence · Technische Universität Dresden

   Affiliation as printed

   ScaDS.AI, Dresden, Germany

   TU Dresden, Dresden, Germany