Learning Bayesian belief networks based on the minimum description length principle: Basic properties

被引：0
|
作者：
Suzuki, J [1 ]
机构：
[1] Osaka Univ, Fac Sci, Osaka 5600043, Japan
来源：
IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES | 1999年 / E82A卷 / 10期
关键词：
minimum description length principle; Bayesian belief network; Chow and Liu algorithm; Cooper and Herskovits procedure; MDL-based procedure; stochastic rule learning;
D O I：
暂无
中图分类号：
TP3 [计算技术、计算机技术];
学科分类号：
0812 ;
摘要：
This paper addresses the problem of learning Bayesian belief networks (BBN) based on the minimum description length (MDL) principle. First, we give a formula of description length based on which the MDL-based procedure learns a BBN. Secondly, we point out that the difference between the MDL-based and Cooper and Herskovits procedures is essentially in the priors rather than in the approaches (MDL and Bayesian), and recommend a class of priors from which the formula is obtained. Finally, we show as a merit of using the formula that a modified version of the Chow and Liu algorithm is obtained. The modified algorithm finds a set of trees rather than a spanning tree based on the MDL principle.
引用
收藏
页码：2237 / 2245
页数：9
相关论文
共 50 条
  • [31] The minimum description length principle for pattern mining: a survey
    Galbrun, Esther
    DATA MINING AND KNOWLEDGE DISCOVERY, 2022, 36 (05) : 1679 - 1727
  • [32] The minimum description length principle for pattern mining: a survey
    Esther Galbrun
    Data Mining and Knowledge Discovery, 2022, 36 : 1679 - 1727
  • [33] Optimal work extraction and the minimum description length principle
    Touzo, Leo
    Marsili, Matteo
    Merhav, Neri
    Roldan, Edgar
    JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT, 2020, 2020 (09):
  • [34] Integrating inductive learning to the ripple down rules method with the minimum description length principle
    Wada, T., 2001, Japanese Society for Artificial Intelligence, OS Bldg #402, Tsukudo-cho, Shinjuku-ku, Tokyo, 162-082, Japan (16)
  • [35] Adaptive ripple down rules method based on minimum description length principle
    Yoshida, T
    Wada, T
    Motoda, H
    Washio, T
    2002 IEEE INTERNATIONAL CONFERENCE ON DATA MINING, PROCEEDINGS, 2002, : 530 - 537
  • [36] Minimum Description Length Recurrent Neural Networks
    Lan, Nur
    Geyer, Michal
    Chemla, Emmanuel
    Katzir, Roni
    TRANSACTIONS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, 2022, 10 : 785 - 799
  • [37] Information Geometry and Minimum Description Length Networks
    Sun, Ke
    Wang, Jun
    Kalousis, Alexandros
    Marchand-Maillet, Stephane
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 37, 2015, 37 : 49 - 58
  • [38] Inference of Gene Regulatory Networks Based on a Universal Minimum Description Length
    Dougherty, John
    Tabus, Ioan
    Astola, Jaakko
    EURASIP JOURNAL ON BIOINFORMATICS AND SYSTEMS BIOLOGY, 2008, (01)
  • [39] UNSUPERVISED HANDWRITTEN GRAPHICAL SYMBOL LEARNING Using Minimum Description Length Principle on Relational Graph
    Li, Jinpeng
    Mouchere, Harold
    Viard-Gaudin, Christian
    KDIR 2011: PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND INFORMATION RETRIEVAL, 2011, : 172 - 178
  • [40] Pruning fuzzy ARTMAP using the minimum description length principle in learning from clinical databases
    Lin, TH
    Soo, VW
    NINTH IEEE INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE, PROCEEDINGS, 1997, : 396 - 403
12345