Novelty-organizing team of classifiers - A team-individual multi-objective approach to reinforcement learning

研究成果: Chapter in Book/Report/Conference proceedingConference contribution

6 被引用数 (Scopus)

抄録

In reinforcement learning, there are basically two spaces to search: value-function space and policy space. Consequently, there are two fitness functions each with their associated trade-offs. However, the problem is still perceived as a single-objective one. Here a multi-objective reinforcement learning algorithm is proposed with a structured novelty map population evolving feedforward neural models. It outperforms a gradient based continuous input-output state-of-art algorithm in two problems. Contrary to the gradient based algorithm, the proposed one solves both problems with the same parameters and smaller variance of results. Moreover, the results are comparable even with other discrete action algorithms of the literature as well as neuroevolution methods such as NEAT. The proposed method brings also the novelty map population concept, i.e., a novelty map-based population which is less sensitive to the input distribution and therefore more suitable to create the state space. In fact, the novelty map framework is shown to be less dynamic and more resource efficient than variants of the self-organizing map.

本文言語英語
ホスト出版物のタイトルProceedings of the SICE Annual Conference
出版社Society of Instrument and Control Engineers (SICE)
ページ1785-1792
ページ数8
ISBN(電子版)9784907764463
DOI
出版ステータス出版済み - 10 23 2014
イベント2014 53rd Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2014 - Sapporo, 日本
継続期間: 9 9 20149 12 2014

出版物シリーズ

名前Proceedings of the SICE Annual Conference

その他

その他2014 53rd Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2014
Country日本
CitySapporo
Period9/9/149/12/14

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Computer Science Applications
  • Electrical and Electronic Engineering

フィンガープリント 「Novelty-organizing team of classifiers - A team-individual multi-objective approach to reinforcement learning」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル