Presentation 2010-11-19
Learning Goal-Directed Reasoning by Relational Reinforcement Learning
Hideaki ITOH, Hisao FUKUMOTO, Hiroshi WAKUYA, Tatsuya FURUKAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Goal-directed reasoning is a type of reasoning in which an agent, who does not know how to achieve its goal, infers the way to achieve the goal by setting an achievable subgoal from which the ultimate goal can be achieved. This type of reasoning is simple yet useful in many problem-solving situations. In this paper, we use a relational reinforcement learning method, which is a combination of predicate logic and reinforcement learning, to make an agent who learns through experience the algorithm to perform the goal-directed reasoning. Especially, we report some difficulty in the use of gradient methods when learning an algorithm that involves storing some information into memory and using the stored information after a period of time.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) goal-directed reasoning / relational reinforcement learning / predicate logic
Paper # AI2010-31
Date of Issue

Conference Information
Committee AI
Conference Date 2010/11/12(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 Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Learning Goal-Directed Reasoning by Relational Reinforcement Learning
Sub Title (in English)
Keyword(1) goal-directed reasoning
Keyword(2) relational reinforcement learning
Keyword(3) predicate logic
1st Author's Name Hideaki ITOH
1st Author's Affiliation Graduate School of Science and Engineering, Saga University()
2nd Author's Name Hisao FUKUMOTO
2nd Author's Affiliation Graduate School of Science and Engineering, Saga University
3rd Author's Name Hiroshi WAKUYA
3rd Author's Affiliation Graduate School of Science and Engineering, Saga University
4th Author's Name Tatsuya FURUKAWA
4th Author's Affiliation Graduate School of Science and Engineering, Saga University
Date 2010-11-19
Paper # AI2010-31
Volume (vol) vol.110
Number (no) 301
Page pp.pp.-
#Pages 6
Date of Issue