Decision-theoretic planning with non-Markovian rewards

Thiebaux, Sylvie M; Gretton, Charles; Slaney, John K; Price, David; Kabanza, Froduald

Decision-theoretic planning with non-Markovian rewards

Date

2006

Authors

Thiebaux, Sylvie M

Gretton, Charles

Slaney, John K

Price, David

Kabanza, Froduald

Publisher

Morgan Kauffman Publishers

Abstract

A decision process in which rewards depend on history rather than merely on the current state is called a decision process with non-Markovian rewards (NMRDP). In decision-theoretic planning, where many desirable behaviours are more naturally expressed as

Keywords

Keywords: Computer software; Decision theory; Dynamic programming; Formal logic; Functions; Heuristic methods; Planning; Hand-coded tracks; Markovian decision process (MDP) model; Non-Markovian rewards (NMRDP); Software platform; Artificial intelligence

Citation

Journal of Artificial Intelligence Research 25 (2006): 17-74

URI

http://hdl.handle.net/10440/32
http://digitalcollections.anu.edu.au/handle/10440/32

Collections

ANU Research Publications

Source

Journal of Artificial Intelligence Research

Type

Journal article

Downloads

File

Description

Thiebaux_Decision2006.pdf (682.74 KB)

Full item page

Decision-theoretic planning with non-Markovian rewards

Date

Authors

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Description

Keywords

Citation

URI

Collections

Source

Type

Book Title

Entity type

Access Statement

License Rights

DOI

Restricted until

Downloads