基于属性加权朴素贝叶斯的冠心病辨证模型  

Coronary Heart Disease Diagnosis Model Based on Weighted Bayesian Categorization

作　　者：刘智[1] 桑国明[1] 鲁明羽[1] 

机构地区：[1]大连海事大学信息科学技术学院,辽宁大连116026

出　　处：《广西师范大学学报（自然科学版）》2008年第4期67-70,共4页Journal of Guangxi Normal University:Natural Science Edition

基　　金：国家自然科学基金资助项目(60773084;J0724003);教育部博士点基金资助项目(20070151009)

摘　　要：在原有中医药冠心病临床治疗数据采集系统的基础上,使用中医证型的辨证相关因素,提出属性加权朴素贝叶斯算法,并应用到冠心病中医证型的分类模型之中。实验结果显示,对于冠心病4种证型的分类,运用属性加权朴素贝叶斯分类算法都略高于朴素贝叶斯分类算法。实验结果表明属性加权朴素贝叶斯分类算法在中医冠心病临床诊断中具有良好的分类性能。To make the diagnosis of coronary heart disease more effective, the symptoms of the disease based on Chinese traditional medical clinical diagnosis must be determined. This paper proposes an attributes weighted Naive Bayesian algorithm and describes its usage in the Chinese traditional medical clinical classification model for coronary heart disease. The experimental results show that the attribute weighted Bayesian classification algorithm has better performance than the Naive Bayesian classification algorithm on any classification of the four symptoms of the coronary heart disease, and the attribute weighted Bayesian classification algorithm has better classification capability in Chinese traditional medi- cal clinical diagnosis model.

关 键 词：属性加权朴素贝叶斯 冠心病 属性权重 分类模型 

分 类 号：TP393.04[自动化与计算机技术—计算机应用技术]