Learning Bayesian network classifier based on prediction ability
基于预测能力的贝叶斯网络分类器学习*

This paper analyzed representative Bayesian classifiers, gave concept and estimating methods of prediction ability between variables, presented a new learning method of the Bayesian network classifier based prediction ability, finished experiment by data from UCI. Experimental results show that the structure of the Bayesian network classifier learned by this method is simpler and has better classifying ability.