线性指数危险率模型的贝叶斯判别分析

数学物理学报 ›› 2003, Vol. 23 ›› Issue (4): 436-443.

线性指数危险率模型的贝叶斯判别分析

魏立力, 张文修

西安交通大学理学院信息与系统科学研究所西安 710049

出版日期:2003-08-25 发布日期:2003-08-25
基金资助:
宁夏自然科学基金(1999A002)资助项目

Bayesian Procedure in Classification Problem With Linear Expon ential Hazard Function

WEI Li-Li, ZHANG Wen-Xiu

西安交通大学理学院信息与系统科学研究所西安 710049

Online:2003-08-25 Published:2003-08-25
Supported by:
宁夏自然科学基金(1999A002)资助项目

摘要/Abstract

摘要：

设有两个总体Π0和Π1，其危险率为具有不同参数的线性函数。对于待观测的寿命样本X，给出了相应的判别分析问题的Bayes停止判决法则，其中损失函数包括试验费用和误判损失两部分。


关键词: 危险率函数;判别分析;Bayesian法则;损失函数

Abstract:

Let Π0 and Π1 be two populations, the hazard functions of which are linear exponential functions with different param eters.For the life data X in an experiment, the Bayesian stopping rule and decision r ule of the classification problem are given in this paper, where the loss functi o n of the model is the sum of the cost of the experiment time and the loss due to the wrong decision.


Key words: Hazard function, Classification problem; Loss function, Bayesian procedure

中图分类号:

62C10

魏立力, 张文修. 线性指数危险率模型的贝叶斯判别分析[J]. 数学物理学报, 2003, 23(4): 436-443.

WEI Li-Li, ZHANG Wen-Xiu. Bayesian Procedure in Classification Problem With Linear Expon ential Hazard Function[J]. Acta mathematica scientia,Series A, 2003, 23(4): 436-443.