講演名 2011-03-07
Fusing Learning Strategies to Learn Various Tasks with Single Configuration
,
PDFダウンロードページ PDFダウンロードページへ
抄録(和)
抄録(英) This paper proposes a method to fuse learning strategies (LSs) in reinforcement learning framework. Generally, we need to choose a suitable LS for each task respectively. In contrast, the proposed method automates this selection by fusing LSs. The LSs fused in this paper includes a transfer learning, a hierarchical RL, and a model based RL. The proposed method has a wide applicability. When the method is applied to a motion learning task, such as a crawling task, the performance of motion may be improved compared to an agent with a single LS. The method also can be applied to a navigation task by hierarchically combining already learned motions, such as a crawling and a turning. This paper demonstrates a maze task of a humanoid robot where the robot learns not only a path to goal, but also a crawling and a turning motions.
キーワード(和)
キーワード(英) Learning System / Reinforcement Learning / Modularization / Motion Learning / Humanoid Robot
資料番号 NC2010-154
発行日

研究会情報
研究会 NC
開催期間 2011/2/28(から1日開催)
開催地(和)
開催地(英)
テーマ(和)
テーマ(英)
委員長氏名(和)
委員長氏名(英)
副委員長氏名(和)
副委員長氏名(英)
幹事氏名(和)
幹事氏名(英)
幹事補佐氏名(和)
幹事補佐氏名(英)

講演論文情報詳細
申込み研究会 Neurocomputing (NC)
本文の言語 ENG
タイトル(和)
サブタイトル(和)
タイトル(英) Fusing Learning Strategies to Learn Various Tasks with Single Configuration
サブタイトル(和)
キーワード(1)(和/英) / Learning System
第 1 著者 氏名(和/英) / Akihiko YAMAGUCHI
第 1 著者 所属(和/英)
Graduate School of Information Science, Nara Institute of Science and Technology
発表年月日 2011-03-07
資料番号 NC2010-154
巻番号(vol) vol.110
号番号(no) 461
ページ範囲 pp.-
ページ数 6
発行日