Dynamic Programming Structure Learning Algorithm of Bayesian Network Integrating MWST and Improved MMPC

Table 3

Time consumption and accuracy of five algorithms

n	Running time (s)					Scoring error percentage (%)
	DP	SMDP	MEDP	DPCMB	MMPCDP	DP	SMDP	MEDP	DPCMB	MMPCDP
10	15	2	2	1	2	1.0367e-14	1.6415e-14	1.4829e-13	4.1813e-13	1.0354e-13
11	24	4	4	2	3	2.1546e-12	2.2428e-12	2.7766e-12	2.8182e-12	2.9145e-12
12	99	24	11	5	9	5.1105e-13	1.1126e-12	3.2659e-12	0.3510	2.3863e-12
13	262	115	26	12	12	5.9007e-13	1.0534e-12	1.2339e-12	2.1302e-12	0.6110
14	782	253	60	19	15	1.4377e-13	2.9742e-13	3.0249e-13	0.3085	0.3262
15	3128	638	148	27	18	2.2582e-14	7.0055e-14	7.9981e-14	0.0208	7.0080e-14
16	7658	2534	303	41	23	1.5372e-13	1.7624e-13	1.6863e-13	1.2875e-12	0.2656
17	57770	9326	621	103	38	1.0081e-13	1.0435e-13	1.2154e-12	1.0734e-12	1.0307e-12
18	98492	24200	1635	561	45	2.9685e-14	4.7709e-14	8.0055e-14	0.0163	1.2658e-13
19	—	—	5302	734	67	—	—	1.3837e-12	2.0963e-12	1.7139e-12
20	—	—	18525	5447	103	—	—	1.1327e-13	1.3587e-12	1.2378e-12
21	—	—	93363	10328	153	—	—	1.9672e-14	0.6803	0.6036