东北大学学报(自然科学版) ›› 2021, Vol. 42 ›› Issue (4): 509-515.DOI: 10.12068/j.issn.1005-3026.2021.04.008

• 信息与控制 • 上一篇    下一篇

基于迁移学习的贝叶斯网络参数学习方法

王姝1, 关展旭1, 王晶1, 孙晓辉2   

  1. (1.东北大学 信息科学与工程学院, 辽宁 沈阳110819; 2.大连天籁安全风险管理技术有限公司, 辽宁 大连116021)
  • 修回日期:2020-09-06 接受日期:2020-09-06 发布日期:2021-04-15
  • 通讯作者: 王姝
  • 作者简介:王姝(1979-),女,辽宁沈阳人,东北大学副教授.
  • 基金资助:
    国家自然科学基金资助项目(61973057);矿冶过程自动控制技术国家(北京市)重点实验室开放课题(BGRIMM-KZSKL-2018-09).

Bayesian Network Parameter Learning Method Based on Transfer Learning

WANG Shu1, GUAN Zhan-xu1, WANG Jing1, SUN Xiao-hui2   

  1. 1.School of Information Science & Engineering, Northeastern University, Shenyang 110819, China; 2.Dalian Tianlai Security Risk Management Technology Limited Company, Dalian 116021, China.
  • Revised:2020-09-06 Accepted:2020-09-06 Published:2021-04-15
  • Contact: WANG Shu
  • About author:-
  • Supported by:
    -

摘要: 针对贝叶斯网络参数迁移过程中对源域及目标域限定条件较多等问题,在考虑源域-目标域多种信息形式的情况下,提出一种基于贝叶斯网络参数迁移学习的统一框架.该方法综合考虑了源域结构和数据量在迁移中的作用,在结构相似性的基础上,探讨了备选源域数据量对参数迁移的影响.在迁移过程中引入与目标域数据相关的平衡系数.通过平衡系数将目标域数据与迁移过程联系起来,实现平衡系数的自动调节.Asia网络验证了本文方法的准确性.

关键词: 贝叶斯网络;参数学习;迁移学习;结构相似性;平衡系数

Abstract: In order to solve the problem that there are many restrictions on the source domain and the target domain in the process of Bayesian network parameter transfer, a unified framework based on Bayesian network parameter transfer learning was proposed under the condition of considering multiple information forms of source domain and target domain. The method considers the role of source domain structure and data volume in the migration. On the basis of structural similarity, the influence of alternative source domain data volume on parameter migration was discussed. The balance coefficient related to the target domain data was introduced in the migration process. According to the balance coefficient, the target domain data was linked with the migration process to realize the automatic adjustment of the balance coefficient. The Asia network verifies the accuracy of the method in this paper.

Key words: Bayesian network; parameter learning; transfer learning; structural similarity; equilibrium coefficient

中图分类号: