準エキスパート集団からのアンサンブル逆強化学習

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2017/04/01

タイトル(英語): Ensemble Inverse Reinforcement Learning from Semi-Expert Agents

著者名: 冨田　真司（横浜国立大学大学院工学研究院），濱津　文哉（横浜国立大学大学院工学研究院），濱上　知樹（横浜国立大学大学院工学研究院）

著者名(英語): Shinji Tomita (Graduate School of Engineering, Yokohama National University), Fumiya Hamatsu (Graduate School of Engineering, Yokohama National University), Tomoki Hamagami (Graduate School of Engineering, Yokohama National University)

キーワード: 逆強化学習，強化学習，見習い学習，アンサンブル学習，Adaboost　　inverse reinforcement learning，reinforcement learning，apprenticeship learning，ensemble learning，Adaboost

要約(英語): Ensemble inverse reinforcement learning from semi-experts' behavior is proposed. In many inverse reinforcement learning (IRL) problem, the expert agent which has ideal rewards for achieving the goal is supposed to be existing. However, in real world problem, the expert is not always observed. Moreover, the estimated reward function includes the bias depending on its inherent behavior if the reward for achieving the goal task is estimated from one agent. In order to overcome the limitation of IRL, we apply Adaboost, one of ensemble and boosting approach, to IRL and integrate estimated reward functions from semi-expert agents. To confirm the effectiveness of the proposed method in the grid world including incomplete areas, we compared the results of reinforcement learning using estimated reward functions and integrated reward function by simulation. The simulation result shows the proposed method can estimate the reward adaptively.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.137 No.4 （2017） 特集：量子・情報・エレクトロニクス医療応用

本誌掲載ページ: 667-673 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/137/4/137_667/_article/-char/ja/