Research on Dynamic Programming Strategy of Bayesian Network Structure Learning

Table 7

Accuracy comparison between three algorithms under UCI data sets.

Data sets	Nodes	Sample sizes	Tns	DFSDP	CBDP	BFSDP
monks_T	7	432	19	−2.68e + 003	−2.88e + 003	−2.68e + 003
Car	7	1728	27	−1.36e + 004	−1.36e + 004	−1.36e + 004
Diabetes	9	768	113	−1.22e + 004	1.79 e + 004	−1.22e + 004
Abalone	9	4177	245	OM	OM	OM
Nursery	9	12960	35	−1.26e + 005	−1.26e + 005	−1.26e + 005
Contracep	10	1473	46	−1.53e + 004	−2.47e + 004	−1.53e + 004
Wine	14	178	95	−2.88e + 003	−5.81e + 003	−2.88e + 003
Heart	14	270	82	−4.29e + 003	−7.96e + 003	−4.29e + 003
Australian	15	690	108	OT	OT	−1.11e + 004
Zoo	17	101	42	OT	OT	−638.3446