zbMATH — the first resource for mathematics

Learning Bayesian networks from data by particle swarm optimization. (English) Zbl 1113.68435
Summary: Learning Bayesian network is an NP-hard problem. When the number of variables is large, the process of searching optimal network structure could be very time consuming and tends to return a structure which is local optimal. The Particle Swarm Optimization (PSO) is introduced for the solution of the problem of learning Bayesian networks and a novel structure learning algorithm using PSO is proposed. To search in directed acyclic graphs spaces efficiently, a discrete PSO algorithm especially for structure learning based on the characteristics of Bayesian networks is proposed.

68T05 Learning and adaptive systems in artificial intelligence