A novel method for combining Bayesian networks, theoretical analysis, and its applications

作者：

Highlights：

• A generic method is developed to combine structures and parameters of any Bayesian networks (BNs).

• A theoretical analysis shows the distinctive advantages of the proposed BNs combination method.

• The proposed method is applied in recommender systems, bank direct marketing and disease diagnosis.

摘要

Highlights•A generic method is developed to combine structures and parameters of any Bayesian networks (BNs).•A theoretical analysis shows the distinctive advantages of the proposed BNs combination method.•The proposed method is applied in recommender systems, bank direct marketing and disease diagnosis.

论文关键词：Bayesian networks combination,Conditional independencies,Association degree superpose,Knowledge fusion

论文评审过程：Received 26 July 2013, Revised 4 November 2013, Accepted 10 December 2013, Available online 19 December 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.12.005