A Test for Extreme Value Domain of Attraction

A Test for Extreme Value Domain of Attraction
复制标题

极值吸引力域的检验

DOI：
--
复制
发表时间：
1992
期刊：
影响因子：
--
通讯作者：
Z. Wang
Z. Wang
中科院分区：
--
文献类型：
--
作者：
A. M. Hasofer;Z. Wang

文献摘要

被引文献

相似文献

摘要提出了一个简单的统计量来检验样本来自Gumbel分布吸引域中的一个分布的假设。它基于前k阶统计量，是Shapiro-Wilk拟合优度统计量的推广。从理论上和模拟上研究了检验的临界区域及其对样本来自另一个吸引域中的分布的选择的功效。结果表明，该方法的功效优于以前提出的其他方法，具有上级的功效。
Abstract A simple statistic is proposed to test the hypothesis that a sample comes from a distribution in the domain of attraction of the Gumbel distribution. It is based on the top k order statistics and is a generalization of the Shapiro–Wilk goodness-of-fit statistic. The critical region of the test and its power against the alternative that the sample comes from a distribution in another domain of attraction are studied theoretically and by simulation. The power turns out to be superior to that of other tests previously proposed.