指数分布样本中异常数据检验的有效性

ON EFFECTIVENESS OF TESTING FOR OUTLIERS IN EXPONENTIAL SAMPLES

  • 摘要: 本文讨论指数样本中异常数据检验的有效性,证明了Fisher型统计量在异常值检验中的某种优良性。用随机模拟方法比较了Fisher统计量,Epstoin统计量和Dixon统计量在指数样本异常值检验中的功效。

     

    Abstract: It’s frequently necessary to test whether there exist outliers in exponential samples. In this paper, we’ll prove the Fisher procedure to be in some sense optimum and wealso compare some methods for testing outliers in exponential samples by means of a computer.

     

/

返回文章
返回