程从华.基于指数模型的R = P(Y < X < Z)统计推断[J].海南师范大学学报自科版,2020,33(2):170-178 |
基于指数模型的R = P(Y < X < Z)统计推断 |
Statistical Inference for R = P(Y < X < Z) of Exponential Distribution |
|
DOI:10.12051/j.issn.1674-4942.2020.02.009 |
中文关键词: ROC曲线 指数分布 统计推断 |
英文关键词: ROC curves exponential distribution statistical inference |
基金项目:广东省自然科学基金项(2018A030313829);广东省普通高校特色创新类项目(2019KTSCX202);广东省高等教
育教学研究和改革项目(2019625);肇庆教育发展研究院项目(ZQJYY2019033);肇庆学院青年项目(201930) |
|
摘要点击次数: 610 |
全文下载次数: 227 |
中文摘要: |
文章讨论了接收器工作特性曲线(ROC)指标R = P(Y < X < Z)的统计推断问题,分别
给出了R的不同点估计和区间估计,其中点估计包括最大似然估计和贝叶斯估计,区间估计包括正
态近似区间估计、精确区间估计以及贝叶斯可信度置信区间估计。通过数值模拟,比较了上述估
计方法的优劣,并且给出了选择不同估计方法的建议,数值模拟结果表明上述估计方法是可行的。 |
英文摘要: |
In this article, the statistical inference of the receiver-operating characteristic(ROC) curves R = P(Y < X < Z)
is provided. Different point estimations and interval estimations are proposed. The point estimators obtained include maxi‐
mum likelihood and Bayesian estimators. The interval estimations obtained include approximate intervals, exact confidence
intervals and Bayesian credible intervals. Different methods and the corresponding confidence intervals are compared using
simulation studies, and suggestion is given to choose different estimation method. It shows that these methods are applica‐
ble. |
查看全文
查看/发表评论 下载PDF阅读器 |
关闭 |