Supervised Dimensionality Reduction for the Algorithm Selection Problem

Notice, Danielle and Pavlidis, Nicos and Kheiri, Ahmed (2025) Supervised Dimensionality Reduction for the Algorithm Selection Problem. In: Advances in Computational Intelligence Systems : Contributions Presented at the 23rd UK Workshop on Computational Intelligence (UKCI 2024), September 2-4, 2024, Ulster University, Belfast, UK. Advances in Intelligent Systems and Computing . Springer, Cham, pp. 85-97. ISBN 9783031788567

Text (sup_DR_conf_paper)
sup_DR_conf_paper_Copy_.pdf - Accepted Version
Download (3MB)

Abstract

Instance space analysis extends the algorithm selection framework by enabling the visualisation of problem instances via dimensionality reduction (DR). The lower dimensional projection can also be used as input to predict algorithm performance, or to perform algorithm selection. In this paper we consider two supervised DR methods - partial least squares (PLS) and linear discriminant analysis (LDA) - both as visualisation tools and for the purpose of constructing classification models for algorithm selection. Multinomial logistic regression models are used for the classification problem. We compare PLS and LDA to DR methods previously used in this context on three combinatorial optimisation problems, and show that these methods are as competitive.