The complexity of decentralized control of Markov decision processes

被引:551
作者
Bernstein, DS [1 ]
Givan, R
Immerman, N
Zilberstein, S
机构
[1] Univ Massachusetts, Dept Comp Sci, Amherst, MA 01003 USA
[2] Purdue Univ, Sch Elect & Comp Engn, W Lafayette, IN 47907 USA
关键词
computational complexity; Markov decision process; decentralized control;
D O I
10.1287/moor.27.4.819.297
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 decentralized control of Markov decision processes and give complexity bounds on the worst-case running time for algorithms that find optimal solutions. Generalizations of both the fully observable case and the partially observable case that allow for decentralized control are described. For even two agents, the finite-horizon problems corresponding to both of these models are hard for nondeterministic exponential time. These complexity results illustrate a fundamental difference between centralized and decentralized control of Markov decision processes. In contrast to the problems involving centralized control, the problems we consider provably do not admit polynomial-time algorithms. Furthermore, assuming EXP not equal NEXP, the problems require superexponential time to solve in the worst case.
引用
收藏
页码:819 / 840
页数:22
相关论文
共 50 条
  • [31] Variance minimization of parameterized Markov decision processes
    Xia, Li
    DISCRETE EVENT DYNAMIC SYSTEMS-THEORY AND APPLICATIONS, 2018, 28 (01): : 63 - 81
  • [32] Decision roll and horizon roll processes in infinite horizon discounted Markov decision processes
    White, DJ
    MANAGEMENT SCIENCE, 1996, 42 (01) : 37 - 50
  • [33] Reachability analysis of quantum Markov decision processes
    Ying, Shenggang
    Ying, Mingsheng
    INFORMATION AND COMPUTATION, 2018, 263 : 31 - 51
  • [34] BISIMULATION METRICS FOR CONTINUOUS MARKOV DECISION PROCESSES
    Ferns, Norm
    Panangaden, Prakash
    Precup, Doina
    SIAM JOURNAL ON COMPUTING, 2011, 40 (06) : 1662 - 1714
  • [35] Ranking policies in discrete Markov decision processes
    Dai, Peng
    Goldsmith, Judy
    ANNALS OF MATHEMATICS AND ARTIFICIAL INTELLIGENCE, 2010, 59 (01) : 107 - 123
  • [36] Markov decision processes under observability constraints
    Serin, Y
    Kulkarni, V
    MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 2005, 61 (02) : 311 - 328
  • [37] Complexity of finite-horizon Markov decision process problems
    Mundhenk, M
    Goldsmith, J
    Lusena, C
    Allender, E
    JOURNAL OF THE ACM, 2000, 47 (04) : 681 - 720
  • [38] Affect control processes: Intelligent affective interaction using a partially observable Markov decision process
    Hoey, Jesse
    Schroeder, Tobias
    Alhothali, Areej
    ARTIFICIAL INTELLIGENCE, 2016, 230 : 134 - 172
  • [39] Optimal control of Markov Regenerative Processes
    Pfening, A
    Telek, M
    1998 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-5, 1998, : 663 - 668
  • [40] Aspects of Arranged Marriages and the Theory of Markov Decision Processes
    Amitrajeet a. Batabyal
    Theory and Decision, 1998, 45 : 241 - 253