Selection Of auxiliary objectives in artificial immune systems: Initial explorations

Research output: Chapter in Book/Report/Conference proceedingConference Proceeding (Non-Journal item)

Abstract

Multi-objectivization of artificial immune systems (AIS) is considered for several discrete optimization problems with certain properties. It is performed by adding auxiliary objectives which are selected online using reinforcement learning. The results suggest that using auxiliary objectives may help to improve performance of AIS. It is also shown that reinforcement learning is able to ignore inefficient auxiliary objective in AIS, as it is in EA. To the best of our knowledge, this is the first attempt to study effects of multi-objectivization on AIS.

Original languageEnglish
Title of host publication21st International Conference on Soft Computing
Subtitle of host publicationEvolutionary Computation, Genetic Programming, Swarm Intelligence, Fuzzy Logic, Neural Networks, Chaos, Bayesian Methods, Intelligent Image Processing, Bio-Inspired Robotics, MENDEL 2015
EditorsM. Radek
PublisherBrno University of Technology
Pages47-52
Number of pages6
Volume2015-January
Publication statusPublished - 2015
Externally publishedYes
Event21st International Conference on Soft Computing: Evolutionary Computation, Genetic Programming, Swarm Intelligence, Fuzzy Logic, Neural Networks, Chaos, Bayesian Methods, Intelligent Image Processing, Bio-Inspired Robotics, MENDEL 2015 - Brno, Czech Republic
Duration: 23 Jun 201525 Jun 2015

Publication series

NameMendel
ISSN (Print)1803-3814

Conference

Conference21st International Conference on Soft Computing: Evolutionary Computation, Genetic Programming, Swarm Intelligence, Fuzzy Logic, Neural Networks, Chaos, Bayesian Methods, Intelligent Image Processing, Bio-Inspired Robotics, MENDEL 2015
Country/TerritoryCzech Republic
CityBrno
Period23 Jun 201525 Jun 2015

Keywords

  • BCA
  • Helper-objectives
  • Multi-objectivization
  • Reinforcement learning

Fingerprint

Dive into the research topics of 'Selection Of auxiliary objectives in artificial immune systems: Initial explorations'. Together they form a unique fingerprint.

Cite this