Hierarchical algorithms for discounted and weighted Markov decision processes

被引：7
作者：
Abbad, M
Daoui, C
机构：
[1] Fac Sci Rabat, Rabat, Morocco
[2] Fac Sci & Tech, Beni Mellal, Morocco
来源：
MATHEMATICAL METHODS OF OPERATIONS RESEARCH | 2003年 / 58卷 / 02期
关键词：
discounted MDP; weighted MDP; decomposition; strongly connected classes; graph theory;
D O I：
10.1007/s001860300290
中图分类号：
C93 [管理学]; O22 [运筹学];
学科分类号：
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要：
We consider a discrete time finite Markov decision process (MDP) with the discounted and weighted reward optimality criteria. In [1] the authors considered some decomposition of limiting average MDPs. In this paper, we use an analogous approach for discounted and weighted MDPs. Then, we construct some hierarchical decomposition algorithms for both discounted and weighted MDPs.
引用
收藏
页码：237 / 245
页数：9
相关论文
共 13 条
[1]  
ABBAD M, 2003, IN PRESS OPERATIONS
[2]  
[Anonymous], 1986, STOCHASTIC MODELLING
[3]   A decomposition approach for undiscounted two-person zero-sum stochastic games [J].
Avsar, ZM ;
Baykal-Gürsoy, M .
MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 1999, 49 (03) :483-500
[4]   DISCRETE DYNAMIC-PROGRAMMING [J].
BLACKWELL, D .
ANNALS OF MATHEMATICAL STATISTICS, 1962, 33 (02) :719-&
[5]   MARKOV DECISION-MODELS WITH WEIGHTED DISCOUNTED CRITERIA [J].
FEINBERG, EA ;
SHWARTZ, A .
MATHEMATICS OF OPERATIONS RESEARCH, 1994, 19 (01) :152-168
[6]  
Feinberg EA, 1982, Theory Probability Appl., V27, P486
[7]   COMMUNICATING MDPS - EQUIVALENCE AND LP PROPERTIES [J].
FILAR, JA ;
SCHULTZ, TA .
OPERATIONS RESEARCH LETTERS, 1988, 7 (06) :303-307
[8]  
GONDRAN M, 1990, GRAPHES ALGORITHMES
[9]   A WEIGHTED MARKOV DECISION-PROCESS [J].
KRASS, D ;
FILAR, JA ;
SINHA, SS .
OPERATIONS RESEARCH, 1992, 40 (06) :1180-1187
[10]  
KRASS D, 1989, THESIS J HOPKINS U B
12