Causal constraint pruning for exact learning of Bayesian network structure

• 作者: TAN Xiangyuan GAO Xiaoguang HE Chuchao WANG Zidong
• 来源期刊: 系统工程与电子技术（英文版）
• 年，卷（期）: 2021,（4）
• 所属期刊栏目: SYSTEMS ENGINEERING
• 页码范围: 854-872
• 页数: 19页
• 语种: 英文
• DOI: 10.23919/JSEE.2021.000074

摘要

How to improve the efficiency of exact learning of the Bayesian network structure is a challenging issue. In this paper, four different causal constraints algorithms are added into score calculations to prune possible parent sets, improving state-of-the-art learning algorithms' efficiency. Experimental results indi-cate that exact learning algorithms can significantly improve the efficiency with only a slight loss of accuracy. Under causal con-straints, these exact learning algorithms can prune about 70％possible parent sets and reduce about 60％ running time while only losing no more than 2％ accuracy on average. Additionally, with sufficient samples, exact learning algorithms with causal constraints can also obtain the optimal network. In general, adding max-min parents and children constraints has better res-ults in terms of efficiency and accuracy among these four causal constraints algorithms.