LI Jun-wu, LI Guo-ning, ZHANG Ding
（School of Automation and Electrical Engineering, Lanzhou Jiaotong Uninversity, Lanzhou 730070, China）
Abstract： In view of the shortcomings of traditional Bayesian network (BN) structure learning algorithm, such as low efficiency, premature algorithm and poor learning effect, the intelligent algorithm of cuckoo search (CS) and particle swarm optimization (PSO) is selected. Combined with the characteristics of BN structure, a BN structure learning algorithm of CS-PSO is proposed. Firstly, the CS algorithm is improved from the following three aspects： the maximum spanning tree is used to guide the initialization direction of the CS algorithm, the fitness of the solution is used to adjust the optimization and abandoning process of the solution, and PSO algorithm is used to update the position of the CS algorithm. Secondly, according to the structure characteristics of BN, the CS-PSO algorithm is applied to the structure learning of BN. Finally, chest clinic, credit and car diagnosis classic network are utilized as the simulation model, and the modeling and simulation comparison of greedy algorithm, K2 algorithm, CS algorithm and CS-PSO algorithm are carried out. The results show that the CS-PSO algorithm has fast convergence speed, high convergence accuracy and good stability in the structure learning of BN, and it can get the accurate BN structure model faster and better.
Key words： Bayesian network; structure learning; cuckoo search and particle swarm optimization(CS-PSO)
CLD number： TP183 doi： 10.3969/j.issn.1674-8042.2020.01.012
References
［1］Tien I, Kiureghian A D. Algorithms for Bayesian network modeling and reliability assessment of infrastructure systems. Reliability Engineering & System Safety, 2016, 156(C)： 134-147.
［2］Cai Z, Sun S, Si S, et al. Identifying product failure rate based on a conditional Bayesian network classifier. Expert Systems with Applications, 2011, 38(5)： 5036-5043.
［3］Liang X, Wang H F, Guo J, et al. Bayesian network based fault diagnosis method for on-board equipment of train control system. Journal of the China Railway Society, 2017, 39(8)： 93-100.
［4］Cao Y S, Liu S F, Fang Z G, et al. Generalized grey Bayesian network model for reliability analysis of multi-state system. Systems Engineering and Electronics, 2018, 1(1）： 231-237.
［5］Gu X Q, Wang S T, Ni T G, et al. High interpretative Bayesian Mamdani-Assilan type fuzzy system. Control and Decision, 2018, 33(1）： 16-26.
［6］Ramazzotti D, Nobile M S, Antoniotti M, et al. Structural learning of probabilistic graphical models of cumulative phenomena. In： Proceedings of International Conference on Computational Science, 2018： 678-693.
［7］Guo T, Raymond L. Bayesian network structure learning based on hybrid differential evolution and bee colony algorithm. Pattern Recognition and Artificial Intelligence, 2014, 27(6）： 540-545.
［8］Wang C F, Zhang Y H. Bayesian network structure learning based on unconstrained optimization and genetic algorithm. Control and Decision, 2013, 28(4）： 618-622.
［9］Liu H R, Sun M T, Li L, et al. Study on Bayesian network structure learning algorithm based on ant colony node order optimization. Chinese Journal of Scientific Instrument, 2017, 38(1）： 143-150.
［10］Yang X S, Deb S. Cuckoo search： recent advances and applications. Neural Computing & Applications, 2014, 24(1）： 169-174.
［11］Yang X S, Deb S. Engineering optimization by cuckoo search. International Journal of Mathematical Modelling & Numerical Optimization, 2010, 20(4）： 330-343.
［12］Shi Y H, Eberhart R. Parameter selection in particle swarm optimization. In： Proceedings of the 7th Annual Conference on Evolutionary Programming, 1998： 591-600.
［13］Gao X G, Di R H, Guo Z G. Bayesian network structure learning based on improved particle swarm optimization. Journal of Northwestern Polytechnical University, 2014, 32(5）： 749-755.
［14］Feng D K, Ruan Q, Du L M. Binary cuckoo search algorithm. Journal of Computer Applications, 2013, 33(6）： 1566-1570.
CS-PSO算法在贝叶斯网络结构学习中的应用
李俊武， 李国宁， 张 钉
（兰州交通大学 自动化与电气工程学院， 甘肃 兰州 730070）
摘 要： 针对传统的贝叶斯网络(Bayesian network, BN)结构学习算法运行效率低、 算法易早熟、 学习效果不理想等缺点， 选取布谷鸟(Cuckoo search, CS)和粒子群(Particle swarm optimization, PSO)智能算法， 结合BN结构特点， 提出了一种CS-PSO的BN结构学习算法。 首先， 对CS算法从以下三个方面进行改进： 利用最大支撑树来指导CS算法的初始化方向， 利用解的适应度来调节解的寻优及舍弃过程， 利用PSO算法来进行CS算法的位置更新。 其次根据BN的结构特征， 将CS-PSO算法应用于BN的结构学习。 最后采用chest clinic、 credit和car diagnosis三种经典网络作为仿真模型， 进行贪婪算法、 K2算法、 CS算法和CS-PSO算法的建模和仿真比较。 结果表明， CS-PSO算法在BN的结构学习中， 收敛速度快、 收敛精度高且稳定性好， 可以更快、 更优地得到精确的贝叶斯网络结构模型。
关键词： 贝叶斯网络； 结构学习； CS-PSO算法
引用格式： LI Jun-wu, LI Guo-ning, ZHANG Ding. Application of CS-PSO algorithm in Bayesian network structure learning. Journal of Measurement Science and Instrumentation, 2020, 11（1）： 94-102. ［doi： 10.3969／j.issn.1674-8042.2020.01.012］
[full text view]
