Self-organizing neural models integrating rules and reinforcement learning

Traditional approaches to integrating knowledge into neural network are concerned mainly about supervised learning. This paper presents how a family of self-organizing neural models known as fusion architecture for learning, cognition and navigation (FALCON) can incorporate a priori knowledge and pe...

Full description

Saved in:

Bibliographic Details
Main Authors:	TENG, Teck-Hou, TAN, Zhong-Ming, TAN, Ah-hwee
Format:	text
Language:	English
Published:	Institutional Knowledge at Singapore Management University 2008
Subjects:	self-organizing neural model reinforcement learning supervised learning fusion architecture cognition knowledge refinement symbolic rule temporal-difference learning method Databases and Information Systems OS and Networks
Online Access:	https://ink.library.smu.edu.sg/sis_research/6556 https://ink.library.smu.edu.sg/context/sis_research/article/7559/viewcontent/Integrating_Rules_IJCNN08_av.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!
Institution:	Singapore Management University
Language:	English

id	sg-smu-ink.sis_research-7559
record_format	dspace
spelling	sg-smu-ink.sis_research-75592022-01-10T03:36:23Z Self-organizing neural models integrating rules and reinforcement learning TENG, Teck-Hou TAN, Zhong-Ming TAN, Ah-hwee Traditional approaches to integrating knowledge into neural network are concerned mainly about supervised learning. This paper presents how a family of self-organizing neural models known as fusion architecture for learning, cognition and navigation (FALCON) can incorporate a priori knowledge and perform knowledge refinement and expansion through reinforcement learning. Symbolic rules are formulated based on pre-existing know-how and inserted into FALCON as a priori knowledge. The availability of knowledge enables FALCON to start performing earlier in the initial learning trials. Through a temporal-difference (TD) learning method, the inserted rules can be refined and expanded according to the evaluative feedback signals received from the environment. Our experimental results based on a minefield navigation task have shown that FALCON is able to learn much faster and attain a higher level of performance earlier when inserted with the appropriate a priori knowledge. 2008-06-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/6556 info:doi/10.1109/IJCNN.2008.4634340 https://ink.library.smu.edu.sg/context/sis_research/article/7559/viewcontent/Integrating_Rules_IJCNN08_av.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University self-organizing neural model reinforcement learning supervised learning fusion architecture cognition knowledge refinement symbolic rule temporal-difference learning method Databases and Information Systems OS and Networks
institution	Singapore Management University
building	SMU Libraries
continent	Asia
country	Singapore Singapore
content_provider	SMU Libraries
collection	InK@SMU
language	English
topic	self-organizing neural model reinforcement learning supervised learning fusion architecture cognition knowledge refinement symbolic rule temporal-difference learning method Databases and Information Systems OS and Networks
spellingShingle	self-organizing neural model reinforcement learning supervised learning fusion architecture cognition knowledge refinement symbolic rule temporal-difference learning method Databases and Information Systems OS and Networks TENG, Teck-Hou TAN, Zhong-Ming TAN, Ah-hwee Self-organizing neural models integrating rules and reinforcement learning
description	Traditional approaches to integrating knowledge into neural network are concerned mainly about supervised learning. This paper presents how a family of self-organizing neural models known as fusion architecture for learning, cognition and navigation (FALCON) can incorporate a priori knowledge and perform knowledge refinement and expansion through reinforcement learning. Symbolic rules are formulated based on pre-existing know-how and inserted into FALCON as a priori knowledge. The availability of knowledge enables FALCON to start performing earlier in the initial learning trials. Through a temporal-difference (TD) learning method, the inserted rules can be refined and expanded according to the evaluative feedback signals received from the environment. Our experimental results based on a minefield navigation task have shown that FALCON is able to learn much faster and attain a higher level of performance earlier when inserted with the appropriate a priori knowledge.
format	text
author	TENG, Teck-Hou TAN, Zhong-Ming TAN, Ah-hwee
author_facet	TENG, Teck-Hou TAN, Zhong-Ming TAN, Ah-hwee
author_sort	TENG, Teck-Hou
title	Self-organizing neural models integrating rules and reinforcement learning
title_short	Self-organizing neural models integrating rules and reinforcement learning
title_full	Self-organizing neural models integrating rules and reinforcement learning
title_fullStr	Self-organizing neural models integrating rules and reinforcement learning
title_full_unstemmed	Self-organizing neural models integrating rules and reinforcement learning
title_sort	self-organizing neural models integrating rules and reinforcement learning
publisher	Institutional Knowledge at Singapore Management University
publishDate	2008
url	https://ink.library.smu.edu.sg/sis_research/6556 https://ink.library.smu.edu.sg/context/sis_research/article/7559/viewcontent/Integrating_Rules_IJCNN08_av.pdf
_version_	1770575987575095296

Self-organizing neural models integrating rules and reinforcement learning

Similar Items