A Decomposition Algorithm for Learning Bayesian Networks Based on Scoring Function

Table 1

Results relative to Algorithm 1 with 𝛼=0.01 and 𝛼=0.05: missing edges, extra edges, reversed edges, and computation time.

Algorithm (level 𝛼)	𝑛=500	𝑛=1000	𝑛=2000	𝑛=5000	𝑛=10000
	(2.4, 3.1, 1.2, 6.9951)	(1.6, 3.0, 1.1, 7.6500)	(1.2, 3.0, 1.0, 9.9092)	(0.8, 2.9, 0.9, 12.7819)	(0.6, 2.7, 0.7, 18.8523)
Algorithm 1 (0.01)	(1.0, 1.0, 1.0, 1.0000)	(1.0, 1.0, 1.0, 1.0000)	(1.0, 1.0, 1.0, 1.0000)	(1.0, 1.0, 1.0, 1.0000)	(1.0, 1.0, 1.0, 1.0000)
Algorithm 1 (0.05)	(0.9, 2.0, 1.3, 0.9199)	(0.9, 1.8, 1.3, 0.9330)	(1.0, 1.4, 1.2, 1.1031)	(1.1, 1.1, 1.2, 1.1730)	(1.2, 1.0, 1.1, 1.1883)
K2 algorithm	(1.4, 2.2, 1.3, 1.5819)	(1.3, 1.9, 1.2, 1.7235)	(1.2, 1.5, 1.2, 2.1592)	(1.2, 1.4, 1.1, 2.7847)	(1.0, 1.2, 1.0, 2.9670)