论文:2014,Vol:32,Issue(5):749-755
引用本文:
高晓光, 邸若海, 郭志高. 基于改进粒子群优化算法的贝叶斯网络结构学习[J]. 西北工业大学
Gao Xiaoguang, Di Ruohai, Guo Zhigao. Bayesian Network Structure Learning Based on Improved Particle Swarm Optimization[J]. Northwestern polytechnical university

基于改进粒子群优化算法的贝叶斯网络结构学习
高晓光, 邸若海, 郭志高
西北工业大学 电子信息学院, 陕西 西安 710129
摘要:
贝叶斯网络结构学习是数据挖掘和知识发现领域的重要研究技术之一,在网络结构的搜索空间较大的情况下,传统的二值粒子群优化算法往往存在收敛速度慢,容易陷入局部最优,学习精度较差的缺陷。在传统二值粒子群优化算法基础上,利用互信息限制粒子群算法的初始化,缩小算法的搜索空间,同时构建新的进化模型代替原有的进化公式,使得改进后的算法具有更强的寻优能力。采用ASIA网络作为仿真模型,并与原有算法比较,结果表明,改进算法能够在较少的迭代次数下找到较优的解,并且基本没有增加算法的复杂度。
关键词:    贝叶斯网络    数据挖掘    粒子群优化   
Bayesian Network Structure Learning Based on Improved Particle Swarm Optimization
Gao Xiaoguang, Di Ruohai, Guo Zhigao
Department of Electronics Engineering, Northwestern Polytechnical University, Xi'an 710129, China
Abstract:
Bayesian network structure learning is one of the important research techniques in the domain of datamining and knowledge discovery,when the search space of the network structure is bigger,traditional binary parti-cle algorithms often have some defects such as low convergent speed,falling easily into local optimum and low pre-cision.We improve the classic binary particle swarm optimization algorithm in two respects: particle initializationand update process; the improved algorithm has stronger optimization ability.We compare the proposed algorithmwith the original algorithm using the ASIA network.The results and their analysis show preliminarily that the pro-posed algorithm is able to find the better solution with less number of iterations,without increasing the complexitybasically.
Key words:    Bayesian networks    data mining    particle swarm optimization (PSO)   
收稿日期: 2014-04-02     修回日期:
DOI:
基金项目: 全国高校博士点基金(20116102110026)资助
通讯作者:     Email:
作者简介: 高晓光(1957-),女,西北工业大学教授,主要从事复杂系统建模及效能评估、贝时斯理论研究。
相关功能
PDF(402KB) Free
打印本文
把本文推荐给朋友
作者相关文章
高晓光  在本刊中的所有文章
邸若海  在本刊中的所有文章
郭志高  在本刊中的所有文章

参考文献:
[1] 辛玉林,杜彬彬,徐世友. 基于随机模糊贝叶斯网络的敌我属性融合算法[J]. 控制与决策,2011,26(3): 443-447Xin Yulin,Du Binbin,Xu Shiyou. Friend-or-Foe Fusion Identification Algorithm Based on Bayesian Network Using RandomFuzzy Theory[J]. Control and Decision,2011,26(3): 443-447 (in Chinese)
[2] 陈海洋,高晓光,樊昊. 变结构 DDBNS 的推理算法与多目标识别 J]. 航空学报,2010,31(11): 2222-2227Chen Haiyang,Gao Xiaoguang,Fan Hao. Inference Algorithm of Variable Structure DDBNs and Multi-Target Recognition[J].Acta Aeronautica et Astronautica Sinica,2010,31(11): 2222-2227 (in Chinese)
[3] Pietquin O,Dutoit T. A Probabilitic Framework for Dialog Simulation and Optimal Strategy Learning[J]. IEEE Trans on Speechand Audio Processing,2005,14(2): 589-599
[4] Kennedy J,Eberhart R C. Particle Swarm Opimization[C]∥IEEE International Conference on Neural Network,1995:1942-1948
[5] Kennedy J,Eberhart R C. A Discrete Binary Version of the Particle Swarm Algorithm[C]∥Proc IEEE Int Conf System ManCybericle,1997: 4104-4109
[6] Pampara G,Franken N,Engelbrecht A P. Combining Particle Swarm Optimization with Angle Modulation to Solve Binary Prob-lems[C]∥IEEE Congress Evolution Computation,2005: 89-96
[7] Heng X C,Qin Z,Wang X H,Shao L P. Research on Learning Bayesian Networks by Particle Swarm Optimization[J]. Infor-mation Technology,2006,5(3): 540-545
[8] Heng X C,Qin Z,Tian L,Shao L P. Learning Bayesian Network Structures with Discrete Particle Swarm Optimization Algo-rithm[C]∥IEEE Symposium on Foundation of Computational Intelligence 2007:47-52
[9] Heng X C,Qin Z,Tian L,Shao L P. Research on Structure Learning of Dynamic Bayesian Networks by Particle Swarm Optimi-zation[C]∥IEEE Symposium on Artificial Life,2007: 85-91
[10] 邸若海,高晓光. 基于限制型粒子群优化的贝叶斯网络结构学习[J]. 系统工程与电子技术,2011,33(11): 2423-2427Di Ruohai,Gao Xiaoguang. Bayesian Network Structure Learning Based on Restricted Particle Swarm Optimization[J]. SystemsEngineering and Electronics,2011,33(11): 2423-2427 (in Chinese)
[11] 胡旺,李志蜀. -种更简化而高效的粒子群优化算法[J]. 软件学报,2007(4):861-868Hu Wang,Li Zhishu. A Simpler and More Effective Particle Swarm Optimization Algorithm[J]. Journal of Software,2007(4):861-868 (in Chinese)