基于R软件的Bayes判别分析在医学中的应用

The Medicinal Application of Bayes Discriminant Analysis Based on R

Bayes判别在进行判别分析时考虑到各总体出现的先验概率、预报的先验概率及错判造成的损失,其判别效能优于其他判别方法.对Bayes判别方法详细介绍的基础上,利用R软件对一组舒张压和胆固醇数据分别进行Bayes判别分析、Fisher判别分析和基于距离的判别分析,对比三种不同方法下得到的判别结果,结果表明Bayes判别分析得到的分类结果精度较高,Bayes判别分析在医学领域有较好的应用前景.

Bayes discriminant analysis takes the overall appearance of the prior probability, the forecasting of the prior probability and the loss caused by mistake into account. Its efficiency is better than the other methods. This article does Bayes discriminant analysis, Fisher discriminant analysis and discriminant analysis based on distance on a set of diastolic blood pressure and cholesterol data. Compare the results which derived from the above three different methods. The results show that the precision of classification results based on Bayes discriminant analysis is higher than the others. And the Bayes discriminant analysis has good application prospect in the field of medicine.