高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于改进鲸鱼优化策略的贝叶斯网络结构学习算法

刘浩然 张力悦 范瑞星 王海羽 张春兰

刘浩然, 张力悦, 范瑞星, 王海羽, 张春兰. 基于改进鲸鱼优化策略的贝叶斯网络结构学习算法[J]. 电子与信息学报, 2019, 41(6): 1434-1441. doi: 10.11999/JEIT180653
引用本文: 刘浩然, 张力悦, 范瑞星, 王海羽, 张春兰. 基于改进鲸鱼优化策略的贝叶斯网络结构学习算法[J]. 电子与信息学报, 2019, 41(6): 1434-1441. doi: 10.11999/JEIT180653
Haoran LIU, Liyue ZHANG, Ruixing FAN, Haiyu WANG, Chunlan ZHANG. Bayesian Network Structure Learning Based on Improved Whale Optimization Strategy[J]. Journal of Electronics & Information Technology, 2019, 41(6): 1434-1441. doi: 10.11999/JEIT180653
Citation: Haoran LIU, Liyue ZHANG, Ruixing FAN, Haiyu WANG, Chunlan ZHANG. Bayesian Network Structure Learning Based on Improved Whale Optimization Strategy[J]. Journal of Electronics & Information Technology, 2019, 41(6): 1434-1441. doi: 10.11999/JEIT180653

基于改进鲸鱼优化策略的贝叶斯网络结构学习算法

doi: 10.11999/JEIT180653
基金项目: 国家自然科学基金(51641609)
详细信息
    作者简介:

    刘浩然:男,1980年生,教授,博士生导师,研究方向为无线传感器网络、工业故障检测及预测

    张力悦:男,1994年生,硕士生,研究方向为贝叶斯网络、工业故障检测及预测

    范瑞星:男,1993年生,硕士生,研究方向为贝叶斯网络、工业故障检测及预测

    王海羽:男,1993年生,硕士生,研究方向为群智能算法、贝叶斯网络、工业故障检测及预测

    张春兰:女,1992年生,硕士生,研究方向为工业故障检测及预测

    通讯作者:

    刘浩然 liu.haoran@ysu.edu.cn

  • 中图分类号: TP18

Bayesian Network Structure Learning Based on Improved Whale Optimization Strategy

Funds: The National Natural Science Foundation of China (51641609)
  • 摘要: 针对当前贝叶斯网络结构学习算法易陷入局部最优和寻优效率低的问题,该文提出一种基于改进鲸鱼优化策略的贝叶斯网络结构学习算法。该算法首先提出一种新的方法建立较优的初始种群,然后利用不产生非法结构的交叉变异算子构建适用于贝叶斯网络结构学习的改进捕食行为,同时采用动态调节参数增强算法个体寻优的能力,通过适应度排序更新种群,最终获得最优的贝叶斯网络结构。仿真结果表明,该算法具有全局收敛性,寻优效率高,精确率高于其它同类优化算法。
  • 图  1  ALARM网络中IWOA算法不同措施的精确率对比

    图  2  不同网络中各算法精确率对比

    图  3  不同网络中各算法敏感度对比

    表  1  贝叶斯网络的标准评价量

    评价量物理意义
    TP (True Positive,真正例)得到的网络结构与标准网络结构相同的边的数量
    TN (True Negative,真负例)得到的网络结构与标准结构相同无边情况的数量
    FP (False Positive,假正例)得到的网络结构与标准结构相比增加的边的数量
    FN (False Negative,假负例)得到的网络结构与标准结构相比丢失的边的数量
    Pre (Precision,精确率)真正类别占正类别的比例
    Sen (Sensitivity,敏感度)真正类别占所有类别的比例
    Score网络结构的BIC得分
    t网络结构的执行时间
    下载: 导出CSV

    表  2  不同算法在ASIA网络中的对比

    数据量算法TPTNFPFNt(s)Score
    500IWOA7.047.30.20.96.09–1153.8$ \pm $2.13
    MAK4.244.63.83.43.62–1170.5$ \pm $4.46
    MMHC6.747.21.11.23.54–1157.1$ \pm $6.79
    AESL-GA6.546.31.71.54.21–1165.5$ \pm $9.12
    1000IWOA7.247.600.86.28–2314.8$ \pm $3.52
    MAK4.645.03.43.23.70–2317.5$ \pm $7.47
    MMHC6.945.70.90.84.58–2315.6$ \pm $6.45
    AESL-GA6.446.41.51.66.02–2317.2$ \pm $2.24
    3000IWOA7.447.800.67.81–6711.8$ \pm $5.77
    MAK5.046.23.43.04.32–6723.9$ \pm $5.84
    MMHC7.046.90.70.66.27–6713.3$ \pm $6.46
    AESL-GA6.947.11.20.98.63–6719.5$ \pm $7.31
    5000IWOA7.647.900.49.00–11201.0$ \pm $11.42
    MAK5.046.63.43.05.31–11212.9$ \pm $12.35
    MMHC7.247.50.50.69.01–11208.9$ \pm $15.57
    AESL-GA7.147.71.00.89.56–11210.4$ \pm $14.63
    下载: 导出CSV

    表  3  不同算法在ALARM网络中的对比

    数据量算法TPTNFPFNt(s)Score
    500IWOA37.21271.68.41.4205.72–5575.61$ \pm $7.28
    NOK231.81269.510.15.635.91–5374.76$ \pm $18.05
    MAK32.21261.420.57.2314.16–5580.04$ \pm $10.02
    MMHC27.61244.115.320.22647.78–5598.67$ \pm $20.33
    AESL-GA33.31274.711.312.72647.78–5472.37$ \pm $15.35
    1000IWOA37.61274.27.81.0223.87–10752.25$ \pm $5.35
    NOK232.21270.39.85.437.64–10538.84$ \pm $20.20
    MAK33.41265.219.85.8356.10–10771.08$ \pm $12.36
    MMHC30.21250.410.815.23637.85–10824.35$ \pm $15.89
    AESL-GA35.41275.210.09.23637.85–10964.18$ \pm $8.23
    3000IWOA39.51279.55.60.4476.02–28340.28$ \pm $13.89
    NOK235.61271.97.24.057.35–29043.43$ \pm $14.74
    MAK36.41283.414.24.4505.57–29308.01$ \pm $14.20
    MMHC32.61253.510.212.33963.62–28939.91$ \pm $17.12
    AESL-GA37.01277.16.97.03963.62–29248.78$ \pm $19.30
    5000IWOA40.81280.65.20.2856.10–47893.12$ \pm $17.66
    NOK236.21273.46.53.369.67–48126.57$ \pm $19.58
    MAK37.71270.910.73.7948.41–48259.15$ \pm $20.45
    MMHC35.11277.87.58.05613.55–48677.04$ \pm $19.52
    AESL-GA38.31278.76.34.25613.55–48148.56$ \pm $24.79
    下载: 导出CSV
  • CONTALDI C, VAFAEE F, and NELSON P C. Bayesian network hybrid learning using an elite-guided genetic algorithm[J]. Artificial Intelligence Review, 2018. doi: 10.1007/s10462-018-9615-5
    刘广怡, 李鸥, 宋涛, 等. 基于贝叶斯网络的无线传感网高效数据传输方法[J]. 电子与信息学报, 2016, 38(6): 1362–1367. doi: 10.11999/JEIT151027

    LIU Guangyi, LI Ou, SONG Tao, et al. Energy-efficiency data transmission method in WSN based on Bayesian network[J]. Journal of Electronics &Information Technology, 2016, 38(6): 1362–1367. doi: 10.11999/JEIT151027
    邓歆, 孟洛明. 基于贝叶斯网络的通信网告警相关性和故障诊断模型[J]. 电子与信息学报, 2007, 29(5): 1182–1186.

    DENG Xin and MENG Luoming. Bayesian networks based alarm correlation and fault diagnosis in communication networks[J]. Journal of Electronics &Information Technology, 2007, 29(5): 1182–1186.
    CHICKERING D M. Learning Bayesian Networks is NP-complete[M]. FISHER D and LENZ H J. Learning from Data. New York: Springer, 1996: 121–130.
    SCANAGATTA M, CORANI G, DE CAMPOS C P, et al. Approximate structure learning for large Bayesian networks[J]. Machine Learning, 2018, 107(8/10): 1209–1227. doi: 10.1007/s10994-018-5701-9
    DENNIS D M K, WILLIAMS M R, and SIGMAN M E. Investigative probabilistic inferences of smokeless powder manufacturers utilizing a Bayesian network[J]. Forensic Chemistry, 2017, 3: 41–51. doi: 10.1016/j.forc.2016.12.001
    LIU Hui, ZHOU Shuigeng, LAM W, et al. A new hybrid method for learning Bayesian networks: separation and reunion[J]. Knowledge-Based Systems, 2017, 121: 185–197. doi: 10.1016/j.knosys.2017.01.029
    TSAMARDINOS I, BROWN L E, and ALIFERIS C F. The max-min hill-climbing Bayesian network structure learning algorithm[J]. Machine Learning, 2006, 65(1): 31–78. doi: 10.1007/s10994-006-6889-7
    刘浩然, 孙美婷, 李雷, 等. 基于蚁群节点寻优的贝叶斯网络结构算法研究[J]. 仪器仪表学报, 2017, 38(1): 143–150. doi: 10.19650/j.cnki.cjsi.2017.01.019

    LIU Haoran, SUN Meiting, LI Lei, et al. Study on Bayesian network structure learning algorithm based on ant colony node order optimization[J]. Chinese Journal of Scientific Instrument, 2017, 38(1): 143–150. doi: 10.19650/j.cnki.cjsi.2017.01.019
    刘彬, 王海羽, 孙美婷, 等. 一种通过节点序寻优进行贝叶斯网络结构学习的算法[J]. 电子与信息学报, 2018, 40(5): 1234–1241. doi: 10.11999/JEIT170675

    LIU Bin, WANG Haiyu, SUN Meiting, et al. Learning Bayesian network structure from node ordering searching optimal[J]. Journal of Electronics &Information Technology, 2018, 40(5): 1234–1241. doi: 10.11999/JEIT170675
    MIRJALILI S and LEWIS A. The whale optimization algorithm[J]. Advances in Engineering Software, 2016, 95: 51–67. doi: 10.1016/j.advengsoft.2016.01.008
    MENG Qingfei, CHEN Yuehui, WANG Dong, et al. Learning bayesian networks structure based part mutual information for reconstructing gene regulatory networks[C]. Proceedings of the 13th International Conference on Intelligent Computing, Liverpool, UK, 2017: 647–654. doi: 10.1007/978-3-319-63312-1_57.
    DE CAMPOS C P, SCANAGATTA M, CORANI G, et al. Entropy-based pruning for learning bayesian networks using BIC[J]. Artificial Intelligence, 2018, 260: 42–50. doi: 10.1016/j.artint.2018.04.002
    ZAKHAROV V K and RODIONOV T V. Naturalness of the class of Lebesgue-Borel-Hausdorff measurable functions[J]. Mathematical Notes, 2014, 95(3/4): 500–508. doi: 10.1134/S0001434614030225
    FANG Wei, SUN Jun, CHEN Huanhuan, et al. A decentralized quantum-inspired particle swarm optimization algorithm with cellular structured population[J]. Information Sciences, 2016, 330: 19–48. doi: 10.1016/j.ins.2015.09.055
    陈志敏, 田梦楚, 吴盘龙, 等. 基于蝙蝠算法的粒子滤波法研究[J]. 物理学报, 2017, 66(5): 050502. doi: 10.7498/aps.66.050502

    CHEN Zhimin, TIAN Mengchu, WU Panlong, et al. Intelligent particle filter based on bat algorithm[J]. Acta Physica Sinica, 2017, 66(5): 050502. doi: 10.7498/aps.66.050502
    ADABOR E S, ACQUAAH-MENSAH G K, and ODURO F T. SAGA: A hybrid search algorithm for Bayesian network structure learning of transcriptional regulatory networks[J]. Journal of Biomedical Informatics, 2015, 53: 27–35. doi: 10.1016/j.jbi.2014.08.010
  • 加载中
图(3) / 表(3)
计量
  • 文章访问数:  2570
  • HTML全文浏览量:  916
  • PDF下载量:  80
  • 被引次数: 0
出版历程
  • 收稿日期:  2018-07-03
  • 修回日期:  2019-01-15
  • 网络出版日期:  2019-01-26
  • 刊出日期:  2019-06-01

目录

    /

    返回文章
    返回