Entropy Message Passing

被引：4
|
作者：
Ilic, Velimir M. [1 ]
Stankovic, Miomir S. [2 ]
Todorovic, Branimir T. [1 ]
机构：
[1] Univ Nis, Dept Informat, Fac Sci & Math, Nish, Serbia
[2] Univ Nis, Fac Occupat Safety, Nish, Serbia
来源：
IEEE TRANSACTIONS ON INFORMATION THEORY | 2011年 / 57卷 / 01期
关键词：
Commutative semiring; factor graphs; entropy; expectation maximization; gradient methods; graphical models; sum-product message passing; PARAMETER-ESTIMATION; PROPAGATION; MODELS;
D O I：
10.1109/TIT.2010.2090235
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
The paper proposes a new message passing algorithm for cycle-free factor graphs. The proposed "entropy message passing" (EMP) algorithm may be viewed as sum-product message passing over the entropy semiring, which has previously appeared in automata theory. The primary use of EMP is to compute the entropy of a model. However, EMP can also be used to compute expressions that appear in expectation maximization and in gradient-descent algorithms.
引用
收藏
页码：375 / 380
页数：6
相关论文
共 50 条
  • [1] Accelerated Message Passing for Entropy-Regularized MAP Inference
    Lee, Jonathan N.
    Pacchiano, Aldo
    Bartlett, Peter
    Jordan, Michael, I
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 119, 2020, 119
  • [2] Graph Open-Set Recognition via Entropy Message Passing
    Yang, Lina
    Lu, Bin
    Gan, Xiaoying
    [J]. 23RD IEEE INTERNATIONAL CONFERENCE ON DATA MINING, ICDM 2023, 2023, : 1469 - 1474
  • [3] Passing on the message
    不详
    [J]. VETERINARY RECORD, 2009, 164 (23) : 707 - 707
  • [4] PASSING THE MESSAGE
    CHERRY, M
    [J]. SPAN, 1987, 30 (02): : 55 - 57
  • [5] Are PROMs passing the message?
    Gil-Gouveia, R.
    Oliveira, A.
    [J]. EUROPEAN JOURNAL OF NEUROLOGY, 2023, 30 : 482 - 482
  • [6] Equations for message passing
    Czaja, Ludwik
    [J]. FUNDAMENTA INFORMATICAE, 2006, 72 (1-3) : 81 - 93
  • [7] Convergence Rates of Smooth Message Passing with Rounding in Entropy-Regularized MAP Inference
    Lee, Jonathan N.
    Pacchiano, Aldo
    Jordan, Michael I.
    [J]. INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 108, 2020, 108 : 3003 - 3013
  • [8] MESSAGE PASSING IS NO ANSWER
    SWEAZEY, P
    [J]. ELECTRONIC PRODUCTS MAGAZINE, 1986, 29 (06): : 11 - 11
  • [9] Customized message passing
    Islam, N
    [J]. JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING, 1997, 41 (02) : 205 - 224
  • [10] Variational message passing
    Winn, J
    Bishop, CM
    [J]. JOURNAL OF MACHINE LEARNING RESEARCH, 2005, 6 : 661 - 694
12345