Setup of fuzzy hybrid particle swarms

Research output: Contribution in Book/Catalog/Report/Conference proceedingChapter (peer-reviewed)peer-review

6 Downloads (Pure)

Abstract

This paper presents a framework for systematically investigating and designing fuzzy rulesets for Adaptive Fuzzy Particle Swarm Optimization (AFPSO) algorithms. Training is achieved through Gaussian Process (GP) supported by Gradient Boosted Regression Trees (GBRT). Meta-objective was defined by ranks on various benchmark functions. Validation benchmarks were also performed on GECCO ’20 bound-constrained optimization competition. The resulting variants, particularly those controlling hybridization with Quantum Particle Swarm Optimization (QPSO) surpassed classical Particle Swarm Optimization (PSO), Genetic Algorithm (GA) and Differential Evolution (DE) on the training functions. Some level of generalization was also observed on most of the validation set.
Original languageEnglish
Title of host publicationGECCO '21: Proceedings of the Genetic and Evolutionary Computation Conference Companion
Place of PublicationNew York
PublisherACM Press
Pages207-208
ISBN (Print)978-1-4503-8351-6
DOIs
Publication statusPublished - 2021
EventGECCO '21: Genetic and Evolutionary Computation Conference - Lille, France
Duration: 10 Jul 202114 Jul 2021
Conference number: 21

Conference

ConferenceGECCO '21: Genetic and Evolutionary Computation Conference
Abbreviated titleGECCO
CountryFrance
CityLille
Period10/07/2114/07/21

Keywords

  • particle swarm optimization
  • meta-heuristics

Fingerprint

Dive into the research topics of 'Setup of fuzzy hybrid particle swarms'. Together they form a unique fingerprint.

Cite this