dc.contributor.author | Jacob, D. | |
dc.contributor.author | Polani, D. | |
dc.contributor.author | Nehaniv, C.L. | |
dc.date.accessioned | 2008-03-04T12:10:25Z | |
dc.date.available | 2008-03-04T12:10:25Z | |
dc.date.issued | 2005 | |
dc.identifier.citation | Jacob , D , Polani , D & Nehaniv , C L 2005 , ' Inferring dependencies in Embodiment-based modular reinforcement learning ' , TAROS , vol. 2005 , pp. 103-110 . | |
dc.identifier.other | dspace: 2299/1728 | |
dc.identifier.other | ORCID: /0000-0002-3233-5847/work/86098077 | |
dc.identifier.uri | http://hdl.handle.net/2299/1728 | |
dc.description.abstract | The state-spaces needed to describe realistic--physical embodied agents are extremely large, which presents a serious challenge to classical einforcement learning schemes. In previous work--(Jacob et al., 2005a, Jacob et al., 2005b) we introduced--our EMBER (for EMbodiment-Based modulaR) reinforcement learning system, which describes a novel method for decomposing agents into modules based on the agent s embodiment. This modular decomposition factorises the statespace--and dramatically improves performance--in unknown and dynamic environments. However,--while there are great advantages to be gained from a factorised state-space, the question of dependencies cannot be ignored. We present a development of the work reported in (Jacob et al., 2004) which shows, in a simple example, how dependencies may be identified using a heuristic approach. Results show that the--system is able quickly to discover and act upon--dependencies, even where they are neither simple--nor deterministic. | en |
dc.format.extent | 729356 | |
dc.language.iso | eng | |
dc.relation.ispartof | TAROS | |
dc.title | Inferring dependencies in Embodiment-based modular reinforcement learning | en |
dc.contributor.institution | School of Computer Science | |
dc.contributor.institution | Centre for Computer Science and Informatics Research | |
dc.contributor.institution | Department of Computer Science | |
dc.contributor.institution | Adaptive Systems | |
dc.contributor.institution | School of Physics, Engineering & Computer Science | |
dc.contributor.institution | Centre for Future Societies Research | |
dc.contributor.institution | Science & Technology Research Institute | |
dc.description.status | Peer reviewed | |
rioxxterms.type | Journal Article/Review | |
herts.preservation.rarelyaccessed | true | |