A COMPARISON OF 2 STATISTICS FOR DETECTING CLUSTERING IN ONE-DIMENSION

被引：2

作者：

ANDERSON, NH

TITTERINGTON, DM

机构：

[1] Department of Statistics., University of Glasgow., University Gardens., Glasgow

来源：

JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION | 1995年 / 53卷 / 1-2期

基金：

英国工程与自然科学研究理事会;

关键词：

CLUSTERING OF EVENTS; DENSITY ESTIMATION; SCAN STATISTIC; UNIFORM DISTRIBUTION;

D O I：

10.1080/00949659508811699

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

The paper considers two statistics for testing for non-uniformity of events in one dimension, the Scan Statistic and a statistic based on the supremum of a kernel density estimate. Variations on the Scan Statistic are considered and a number of approximations to the critical values are assessed. The Scan Statistic and the density-supremum statistic are compared in a power study, and modifications to the Scan Statistic are described that adapt to a non-uniform null hypothesis.

引用

页码：103 / 125

页数：23

共 25 条

[1]

[Anonymous], 1989, NUMERICAL RECIPES FO

[2] PREVENTING ASTHMA EPIDEMICS DUE TO SOYBEANS BY DUST-CONTROL MEASURES [J].