A novel Bayesian network structure learning algorithm based on Maximal Information Coefficient

Abstract

Greedy Equivalent Search (GES) is an effective algorithm for Bayesian network problem, which searches in the space of graph equivalence classes. However, original GES may easily fall into local optimization trap because of empty initial structure. In this paper, An improved GES method is prosposed. It firstly makes a draft of the real network, based on Maximum Information Coefficient (MIC) and conditional independence tests. After this step, many independent relations can be found. To ensure correctness, then this draft is used to be a seed structure of original GES algorithm. Numerical experiment on four standard networks shows that NEtoGS (the number of graph structure, which is equivalent to the God Standard network) has big improvement. Also, the total of learning time are greatly reduced. Therefore, our improved method can relatively quickly determine the structure graph with highest degree of data matching.