Skip to main navigation Skip to search Skip to main content

Supervised Dimensionality Reduction for the Algorithm Selection Problem

  • Danielle Notice*
  • , Nicos G. Pavlidis
  • , Ahmed Kheiri
  • *Corresponding author for this work

Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

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.
Original languageEnglish
Title of host publicationAdvances in Computational Intelligence Systems
Subtitle of host publicationContributions Presented at the 23rd UK Workshop on Computational Intelligence (UKCI 2024), September 2-4, 2024, Ulster University, Belfast, UK
EditorsHuiru Zheng, David Glass, Maurice Mulvenna, Jun Liu, Hui Wang
Place of PublicationCham
PublisherSpringer Nature Switzerland AG
Pages85-97
Number of pages13
ISBN (Electronic)9783031788574
ISBN (Print)9783031788567
DOIs
Publication statusPublished - 8 Jan 2024
Event 23rd UK Workshop on Computational Intelligence - Belfast, United Kingdom
Duration: 2 Sept 20244 Sept 2024

Publication series

NameAdvances in Computational Intelligence Systems
PublisherSpringer Cham

Workshop

Workshop 23rd UK Workshop on Computational Intelligence
Abbreviated titleUKCI 2024
Country/TerritoryUnited Kingdom
CityBelfast
Period2/09/244/09/24

Keywords

  • supervised dimensionality reduction
  • algorithm selection
  • classification
  • combinatorial optimisation

Fingerprint

Dive into the research topics of 'Supervised Dimensionality Reduction for the Algorithm Selection Problem'. Together they form a unique fingerprint.

Cite this