广义空间模型的异常值检验

金立斌,戴晓文,石磊

系统科学与数学 ›› 2014, Vol. 34 ›› Issue (1) : 106-118.

PDF(450 KB)
PDF(450 KB)
系统科学与数学 ›› 2014, Vol. 34 ›› Issue (1) : 106-118. DOI: 10.12341/jssms12253
论文

广义空间模型的异常值检验

    金立斌,戴晓文,石磊
作者信息 +

OUTLIER DETECTION IN GENERAL SPATIAL MODEL

    JIN Libin ,DAI Xiaowen ,SHI Lei
Author information +
文章历史 +

摘要

研究了广义空间模型中单个异常值检验问题.分别在均值漂移模型和方差加权模型下导出了检验统计量的具体形式,并给出了在两种异常值模型下检验统计量的近似分布.最后, 通过哥伦布市社区犯罪数据证明了方法的有 效性.

Abstract

This paper studies the score test for single outlier in general spatial model. Score test statistics are derived under the mean-shift model and the variance-weight model respectively. The approximate distributions of score statistics are obtained when outlier presents in two outlier models mentioned above. Finally, we illustrate our method using a real data of columbus neighborhood crime.

关键词

  / 广义空间模型 / 均值漂移模型 / 方差加权模型 / 异常值 / 检验.

引用本文

导出引用
金立斌,戴晓文,石磊. 广义空间模型的异常值检验. 系统科学与数学, 2014, 34(1): 106-118. https://doi.org/10.12341/jssms12253
JIN Libin ,DAI Xiaowen ,SHI Lei. OUTLIER DETECTION IN GENERAL SPATIAL MODEL. Journal of Systems Science and Mathematical Sciences, 2014, 34(1): 106-118 https://doi.org/10.12341/jssms12253
中图分类号: 62J20   
PDF(450 KB)

263

Accesses

0

Citation

Detail

段落导航
相关文章

/