A hybrid methodology for learning belief networks: BENEDICT

被引：39

作者：

Acid, S ^{[1
]}

de Campos, LM ^{[1
]}

机构：

[1] Univ Granada, ETSI Informat, Dept Ciencias Computac & IA, E-18071 Granada, Spain

来源：

INTERNATIONAL JOURNAL OF APPROXIMATE REASONING | 2001年 / 27卷 / 03期

关键词：

belief networks; learning; independence; scoring metrics; minimum d-separating sets;

D O I：

10.1016/S0888-613X(01)00041-X

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Previous algorithms for the construction of belief networks structures from data are mainly based either on independence criteria or on scoring metrics. The aim of this paper is to present a hybrid methodology that is a combination of these two approaches, which benefits from characteristics of each one, and to develop two operative algorithms based on this methodology. Results of the evaluation of the algorithms on the well-known Alarm network are presented, as well as the algorithms performance issues and some open problems. (C) 2001 Elsevier Science Inc, All rights reserved.

引用

页码：235 / 262

页数：28

共 45 条

[21] A new approach for learning belief networks using independence criteria [J].

de Campos, LM ;

Huete, JF .

INTERNATIONAL JOURNAL OF APPROXIMATE REASONING, 2000, 24 (01) :11-37

[22]

deCampos LM, 1997, INT J INTELL SYST, V12, P495, DOI 10.1002/(SICI)1098-111X(199707)12:7<495::AID-INT2>3.0.CO

[23]

2-G

[24]

DECAMPOS LM, 1998, INTELIGENCIA ARTIFIC, V4, P30

[25]

Ford L. R, 1962, FLOWS NETWORKS

[26]

Friedman N, 1996, UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, P252

[27]

Geiger D., 1995, Uncertainty in Artificial Intelligence. Proceedings of the Eleventh Conference (1995), P196

[28]

GEIGER D, 1993, INT J INTELL SYST, V8, P231

[29]

GEIGER D, 1990, 8 NAT C ART INT AAAI, P770

[30]

Glymour C., 1993, LECT NOTES STAT, V1, DOI 10.1007/978-1-4612-2748-9

← 1 2 3 4 5 →