Sample efficiency improvement on neuroevolution via estimation-based elimination strategy

Shengbo Xu, Yuki Inoue, Tetsunari Inamura, Hirotaka Moriguchi, Shinichi Honiden

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we propose estimation-based elimination strategy, which improves sample efficiency of NeuroEvolution (NE) algorithms. The fitness of new individuals was estimated using fitness of individuals evaluated in the past generations. The estimation was achieved by taking average fitness of individuals with high correlation with the new individual. Estimation-based elimination strategy avoids evaluating individuals with low estimated fitness. We adapt estimation-based elimination strategy for state-of-the-art NE algorithms: CMA-NeuroES and CMA-TWEANN. From the experimental results of pole-balancing benchmark tasks, we show that the proposed strategy improves sample efficiency of the NE algorithms.

Original languageEnglish
Title of host publication13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014
PublisherInternational Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)
Pages1537-1538
Number of pages2
ISBN (Electronic)9781634391313
Publication statusPublished - 2014 Jan 1
Event13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014 - Paris, France
Duration: 2014 May 52014 May 9

Publication series

Name13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014
Volume2

Conference

Conference13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014
CountryFrance
CityParis
Period14/5/514/5/9

    Fingerprint

Keywords

  • Evolutionaly computation
  • Fitness estimation
  • Neuroevolution

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Xu, S., Inoue, Y., Inamura, T., Moriguchi, H., & Honiden, S. (2014). Sample efficiency improvement on neuroevolution via estimation-based elimination strategy. In 13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014 (pp. 1537-1538). (13th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2014; Vol. 2). International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS).