Presentation 2012-12-12
The reinforcement learning method that introduced diversive curiosity to induce actions to unknown information.
Takeshi Nakahata, Kazushi Murakoshi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) When the agent is in the face of actions to be learned in reinforcement learning, it is necessary to determine the proper timing of action in order to learn efficiently. Curiosity exists as a motivation which determines subject of interest in such a situation. Curiosity is broadly divided into specific curiosity and diversive curiosity in psychology. In this study, we focused on the diversive curiosity. Diversive curiosity is for directing the interest to a wide range of seeking unknown information in psychology. Based on the definition of diversive curiosiry in psychology, we propose a more efficient learning method which executes a action selection to induce action to unknown information by increasing the action priority according to the number of times to search each piece of information.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Reinforcement Learning / Action Selection / Diversive Curiosity
Paper # NC2012-75
Date of Issue

Conference Information
Committee NC
Conference Date 2012/12/5(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) The reinforcement learning method that introduced diversive curiosity to induce actions to unknown information.
Sub Title (in English)
Keyword(1) Reinforcement Learning
Keyword(2) Action Selection
Keyword(3) Diversive Curiosity
1st Author's Name Takeshi Nakahata
1st Author's Affiliation Toyohashi University of Technology()
2nd Author's Name Kazushi Murakoshi
2nd Author's Affiliation Toyohashi University of Technology
Date 2012-12-12
Paper # NC2012-75
Volume (vol) vol.112
Number (no) 345
Page pp.pp.-
#Pages 6
Date of Issue