Sequential decision making with partially ordered preferences
| Contribuinte(s) |
UNIVERSIDADE DE SÃO PAULO |
|---|---|
| Data(s) |
18/10/2012
18/10/2012
2011
|
| Resumo |
This paper presents new insights and novel algorithms for strategy selection in sequential decision making with partially ordered preferences; that is, where some strategies may be incomparable with respect to expected utility. We assume that incomparability amongst strategies is caused by indeterminacy/imprecision in probability values. We investigate six criteria for consequentialist strategy selection: Gamma-Maximin, Gamma-Maximax, Gamma-Maximix, Interval Dominance, Maximality and E-admissibility. We focus on the popular decision tree and influence diagram representations. Algorithms resort to linear/multilinear programming; we describe implementation and experiments. (C) 2010 Elsevier B.V. All rights reserved. FAPESP[2003/11165-9] FAPESP[2004/09568-0] FAPESP[2005/58090-9] FAPESP[2008/03995-5] |
| Identificador |
ARTIFICIAL INTELLIGENCE, v.175, n.7/Ago, Special Issue, p.1346-1365, 2011 0004-3702 http://producao.usp.br/handle/BDPI/18370 10.1016/j.artint.2010.11.017 |
| Idioma(s) |
eng |
| Publicador |
ELSEVIER SCIENCE BV |
| Relação |
Artificial Intelligence |
| Direitos |
restrictedAccess Copyright ELSEVIER SCIENCE BV |
| Palavras-Chave | #Sequential decision making under uncertainty #Partially ordered preferences #Sets of probability measures #Criteria of choice #Consequentialist and resolute norms #Linear and multilinear programming #IMPRECISE PROBABILITIES #TRANSITION-PROBABILITIES #INFLUENCE DIAGRAMS #EXPECTED UTILITY #CHOICE #UNCERTAINTY #MODELS #BOUNDS #RULES #Computer Science, Artificial Intelligence |
| Tipo |
article original article publishedVersion |
