Title :
A Method for Learning Bayesian Network Structure
Author :
Jingnan Li ; Yingxia Zhang
Author_Institution :
Sch. of Math. & Stat., Xidian Univ., Xian, China
Abstract :
Bayesian network structures from data is an NP-hard problem, In this paper, we propose an approach based on mutual information and PC algorithm methods. This algorithm obain the initial undirected graph using mutual information firstly, obtain a PDAG using PC algorithm. Experimental results show that our method outperforms the PC algorithms under the same conditions, Thus the algorithm decreases the running time and the order of CI tests greatly than the PC algorithm.
Keywords :
belief networks; computational complexity; directed graphs; learning (artificial intelligence); Bayesian network structure learning; CI test; NP-hard problem; PC algorithm methods; PDAG; conditional independence; directed acyclic graphs; mutual information; undirected graph; Algorithm design and analysis; Bayes methods; Boolean functions; Data structures; Mutual information; Probabilistic logic; Skeleton; Bayesian network; conditional indepence test; mutual information; structure learning;
Conference_Titel :
Intelligent Human-Machine Systems and Cybernetics (IHMSC), 2014 Sixth International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4799-4956-4
DOI :
10.1109/IHMSC.2014.156
