Population Markov Chain Monte Carlo

被引:53
|
作者
Laskey, KB
Myers, JW
机构
[1] George Mason Univ, Dept Syst Engn & Operat Res, Fairfax, VA 22030 USA
[2] TRW Co Inc, Reston, VA 20190 USA
关键词
Markov chain Monte Carlo; Metropolis-Hastings algorithm; graphical probabilistic models; Bayesian networks; Bayesian learning; evolutionary algorithms;
D O I
10.1023/A:1020206129842
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Stochastic search algorithms inspired by physical and biological systems are applied to the problem of learning directed graphical probability models in the presence of missing observations and hidden variables. For this class of problems, deterministic search algorithms tend to halt at local optima, requiring random restarts to obtain solutions of acceptable quality. We compare three stochastic search algorithms: a Metropolis-Hastings Sampler (MHS), an Evolutionary Algorithm (EA), and a new hybrid algorithm called Population Markov Chain Monte Carlo, or popMCMC. PopMCMC uses statistical information from a population of MHSs to inform the proposal distributions for individual samplers in the population. Experimental results show that popMCMC and EAs learn more efficiently than the MHS with no information exchange. Populations of MCMC samplers exhibit more diversity than populations evolving according to EAs not satisfying physics-inspired local reversibility conditions.
引用
收藏
页码:175 / 196
页数:22
相关论文
共 50 条
  • [41] OPTIMAL VARIANCE REDUCTION FOR MARKOV CHAIN MONTE CARLO
    Huang, Lu-Jing
    Liao, Yin-Ting
    Chen, Ting-Li
    Hwang, Chii-Ruey
    SIAM JOURNAL ON CONTROL AND OPTIMIZATION, 2018, 56 (04) : 2977 - 2996
  • [42] Regenerative Markov Chain Monte Carlo for Any Distribution
    Minh, Do Le
    Minh, David D. L.
    Nguyen, Andrew L.
    COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 2012, 41 (09) : 1745 - 1760
  • [43] SEQUENTIALLY INTERACTING MARKOV CHAIN MONTE CARLO METHODS
    Brockwell, Anthony
    Del Moral, Pierre
    Doucet, Arnaud
    ANNALS OF STATISTICS, 2010, 38 (06) : 3387 - 3411
  • [44] Markov chain Monte Carlo1. Examples
    Arnab Chakraborty
    Resonance, 2002, 7 (3) : 25 - 34
  • [45] Markov chain Monte Carlo algorithms for the Bayesian analysis of phylogenetic trees
    Larget, B
    Simon, DL
    MOLECULAR BIOLOGY AND EVOLUTION, 1999, 16 (06) : 750 - 759
  • [46] Statistical Robustness of Markov Chain Monte Carlo Accelerators
    Zhang, Xiangyu
    Bashizade, Ramin
    Wang, Yicheng
    Mukherjee, Sayan
    Lebeck, Alvin R.
    ASPLOS XXVI: TWENTY-SIXTH INTERNATIONAL CONFERENCE ON ARCHITECTURAL SUPPORT FOR PROGRAMMING LANGUAGES AND OPERATING SYSTEMS, 2021, : 959 - 974
  • [47] A Markov Chain Monte Carlo Algorithm for Spatial Segmentation
    Raveendran, Nishanthi
    Sofronov, Georgy
    INFORMATION, 2021, 12 (02) : 1 - 14
  • [48] Ensemble preconditioning for Markov chain Monte Carlo simulation
    Benedict Leimkuhler
    Charles Matthews
    Jonathan Weare
    Statistics and Computing, 2018, 28 : 277 - 290
  • [49] Explicit error bounds for Markov chain Monte Carlo
    Rudolf, D.
    DISSERTATIONES MATHEMATICAE, 2012, (485) : 5 - +
  • [50] Segmentation Using Improved Markov Chain Monte Carlo
    Wang, Xiangrong
    Wang, Ming
    Pei, JiaLi
    2011 INTERNATIONAL CONFERENCE ON FUTURE COMPUTER SCIENCE AND APPLICATION (FCSA 2011), VOL 2, 2011, : 172 - 175