BNC-VLA: bayesian network structure learning using a team of variable-action set learning automata

被引:4
作者
Gheisari, S. [1 ]
Meybodi, M. R. [2 ]
Dehghan, M. [2 ]
Ebadzadeh, M. M. [2 ]
机构
[1] Islamic Azad Univ, Dept Comp, Sci & Res Branch, Tehran, Iran
[2] Amirkabir Univ Technol, Comp Engn & Informat Technol Dept, Tehran, Iran
关键词
Bayesian networks; Search and score approach; Structure training; Variable-action set learning automata;
D O I
10.1007/s10489-015-0743-1
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Bayesian Network (BN) is a probabilistic graphical model which describes the joint probability distribution over a set of random variables. One of the most important challenges in the field of BNs is to find an optimal network structure based on an available training dataset. Since the problem of searching the optimal BN structure belongs to the class of NP-hard problems, typically greedy algorithms are used to solve it. In this paper a learning automata-based algorithm has been proposed to solve the BNs structure learning problem. There is a learning automaton corresponding with each random variable and at each stage of the proposed algorithm, named BNC-VLA, a set of learning automata is randomly activated and determined the graph edges that must be appeared in that stage. Finally, the constructed network is evaluated using a scoring function. As BNC-VLA algorithm proceeds, the learning process focuses on the BN structure with higher scores. The convergence of this algorithm is theoretically proved; and also some experiments are designed to evaluate the performance of it. Experimental results show that BNC-VLA is capable of finding the optimal structure of BN in an acceptable execution time; and comparing against other search-based methods, it outperforms them.
引用
收藏
页码:135 / 151
页数:17
相关论文
共 31 条
  • [1] [Anonymous], 2012, Learning automata: an introduction
  • [2] Utilizing distributed learning automata to solve stochastic shortest path problems
    Beigy, Hamid
    Meybodi, M. R.
    [J]. INTERNATIONAL JOURNAL OF UNCERTAINTY FUZZINESS AND KNOWLEDGE-BASED SYSTEMS, 2006, 14 (05) : 591 - 615
  • [3] BEINLICH I.A., 1989, The ALARM monitoring system: A case study with two probabilistic inference techniques for belief networks
  • [4] Cheng J., 1997, Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics, P83
  • [5] Chickering D.M., 1994, MSRTR9414
  • [6] APPROXIMATING DISCRETE PROBABILITY DISTRIBUTIONS WITH DEPENDENCE TREES
    CHOW, CK
    LIU, CN
    [J]. IEEE TRANSACTIONS ON INFORMATION THEORY, 1968, 14 (03) : 462 - +
  • [7] de Campos CP, 2011, J MACH LEARN RES, V12, P663
  • [8] Ant colony optimization for learning Bayesian networks
    de Campos, LM
    Fernández-Luna, JM
    Gámez, JA
    Puerta, JM
    [J]. INTERNATIONAL JOURNAL OF APPROXIMATE REASONING, 2002, 31 (03) : 291 - 311
  • [9] A novel method for combining Bayesian networks, theoretical analysis, and its applications
    Feng, Guang
    Zhang, Jia-Dong
    Liao, Stephen Shaoyi
    [J]. PATTERN RECOGNITION, 2014, 47 (05) : 2057 - 2069
  • [10] Friedman N, 1999, UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS, P206