creators_name: Humphrys, Mark
editors_name: Maes, Pattie
editors_name: Mataric, Maja
editors_name: Meyer, Jean-Arcady
editors_name: Pollack, Jordan
editors_name: Wilson, Stewart W.
type: confpaper
datestamp: 1998-06-09
lastmod: 2011-03-11 08:53:57
metadata_visibility: show
title: Action Selection methods using Reinforcement Learning
ispublished: pub
subjects: bio-ani-behav
subjects: bio-etho
subjects: comp-sci-art-intel
subjects: comp-sci-mach-dynam-sys
subjects: comp-sci-mach-learn
subjects: comp-sci-robot
full_text_status: public
abstract: Action Selection schemes, when translated into precise algorithms, typically involve considerable design effort and tuning of parameters. Little work has been done on solving the problem using learning. This paper compares eight different methods of solving the action selection problem using Reinforcement Learning (learning from rewards). The methods range from centralised and cooperative to decentralised and selfish. They are tested in an artificial world and their performance, memory requirements and reactiveness are compared. Finally, the possibility of more exotic, ecosystem-like decentralised models are considered.
date: 1996
date_type: published
publisher: MIT Press/Bradford Books
pagerange: 135-144
refereed: FALSE
citation:   Humphrys, Mark  (1996) Action Selection methods using Reinforcement Learning.  [Conference Paper]     
document_url: http://cogprints.org/447/2/g.SAB96.ps