Abstract Aiming at the twostage modeling process, a Bayesian network structure refinement method based on conditional mutual information measure is put forward. The detailed procedure of this method is presented, the correctness is proved and the computing complexity is analyzed. A wellknown Bayesian networks-Alarm is applied to show its validity.
[1] Lam W, Nacchus F. Using New Data to Refine a Bayesian Network // Proc of the 10th Conference on Uncertainty in Artificial Intelligence. Seattle, USA, 1994: 383-390 [2] Buntine W. Theory Refinement on Bayesian Networks // Proc of the 7th Conference on Uncertainty in Artificial Intelligence. Los Angeles, USA, 1991: 52-60 [3] Lam W, Bacchus F. Using Causal Information and Local Measures to Learn Bayesian Networks // Proc of the 9th Conference on Uncertainty in Artificial Intelligence. Washington, USA, 1993: 243-250 [4] Pappas A. The Accuracy of a Bayesian Network. Ph.D Dissertation. London, UK: Imperial College. Department of Computing, 2003 [5] Li Gang. Graphical Model for Knowledge Discovery. Ph.D Dissertation. Beijing, China: Chinese Academy of Sciences. Institute of Software, 2001(in Chinese) (李 刚.知识发现的图模型方法.博士学位论文.北京:中国科学院软件研究所, 2001) [6] Friedman N, Goldszmidt M. Sequential Update of Bayesian Network Structure // Proc of the 13th Conference on Uncertainty in Artificial Intelligence. Providence, USA, 1997: 165-174 [7] Bouckaert R R. Properties of Measures for Bayesian Belief Network Learning // Proc of the 10th Conference on Uncertainty in Artificial Intelligence. Seattle, USA, 1994: 102-109 [8] Cheng J, Bell D A, Liu W R. An Algorithm for Bayesian Belief Network Construction from Data // Proc of the Conference on Artificial Intelligence and Statistics. Ft Lauderdale, USA, 1997: 83-90 [9] Pearl J. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. San Mateo, USA: Morgan Kaufmann, 1988 [10] Beinlich I A, Suermondt H J, Chavez R M, et al. The ALARM Monitoring System: A Case Study with Two Probabilistic Inference Techniques for Belief Networks // Proc of the 2nd European Conference on Artificial Intelligence in Medicine. Berlin, Germany: Springer-Verlag, 1989: 247-256 [11]Cooper G F. NESTOR: A Computer-Based Medical Diagnostic Aid That Integrates Causal and Probabilistic Knowledge. Ph.D Dissertation. Stanford, USA: Stanford University. Department of Medical Information Sciences, 1984 [12] Pappas A, Gillies D F. A New Measure for the Accuracy of a Bayesian Network // Proc of the 2nd Mexican International Conference on Artificial Intelligence: Advances in Artificial Intelligence. London, UK: Springer-Verlag, 2002: 411-419 [13] Cheng J, Bell D A, Liu W R. Learning Belief Networks from Data: An Information Theory Based Approach // Proc of the 6th International Conference on Information and Knowledge Management. Las Vegas, USA, 1997: 325-331
2006, Vol. 19 
