Research Article

Upper-Lower Bounds Candidate Sets Searching Algorithm for Bayesian Network Structure Learning

Algorithm 2

LGLA.
(1) Input: Data set D; Variable set ; ;
(2) Initialization: , where ;
(3) For each , If there is a cycle, then call VSTA;
(4) For each , Find makes the highest, ;
(5) If there is another makes is the highest
(6)    If is hold, then ,
(7) End if;
(8) Output as final result.