首页 | 本学科首页   官方微博 | 高级检索  
     检索      


On some robust estimation procedures for quantiles based on data
Authors:Yoeng-Kuan Chang  Deng-Yuan Huang
Institution:  a Department of Business Mathematics, Soochow University, Taipei, Republic of China b Institute of Applied Statistics, Fu-Jen Catholic University, Taipei, Republic of China
Abstract:This paper introduces some robust estimation procedures to estimate quantiles of a continuous random variable based on data, without any other assumptions of probability distribution. We construct a reasonable linear regression model to connect the relationship between a suitable symmetric data transformation and the approximate standard normal statistics. Statistical properties of this linear regression model and its applications are studied, including estimators of quantiles, quartile mean, quartile deviation, correlation coefficient of quantiles and standard errors of these estimators. We give some empirical examples to illustrate the statistical properties and apply our estimators to grouping data.
Keywords:Quantile estimation  symmetric data transformations  quartile mean  quartile deviation  correlation coefficient of quantiles grouping data
本文献已被 InformaWorld 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号