当前位置: X-MOL 学术Natl. Acad. Sci. Lett. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Robust Outlier Detection Method For Multivariate Spatial Data
National Academy Science Letters ( IF 1.2 ) Pub Date : 2021-05-11 , DOI: 10.1007/s40009-021-01056-9
Sweta Shukla , S. Lalitha

Spatial data consist of spatial attributes describing the topology of the object and non-spatial attributes carrying information on the behavioral aspects of the object. This object is termed as a spatial outlier if its non-spatial attributes are significantly different from those in its spatial neighborhood. Here, a robust algorithm based on the Comedian approach is proposed for the detection of spatial outliers in multivariate spatial data. A simulation study is carried for comparisons of the proposed procedure with an existing MCD based spatial outlier detection technique. The simulation results show that the proposed algorithm outperforms the existing algorithm. For demonstrating the effectiveness of the proposed algorithm, its application on real-life CRIME data of India is discussed.



中文翻译:

多元空间数据的鲁棒离群值检测方法

空间数据由描述对象拓扑的空间属性和承载有关对象行为方面信息的非空间属性组成。如果该对象的非空间属性与其空间邻域的属性显着不同,则将该对象称为空间离群值。在此,提出了一种基于Comedian方法的鲁棒算法,用于检测多元空间数据中的空间离群值。进行了仿真研究,以将提出的程序与现有的基于MCD的空间离群值检测技术进行比较。仿真结果表明,该算法优于现有算法。为了证明该算法的有效性,讨论了其在印度真实CRIME数据中的应用。

更新日期:2021-05-11
down
wechat
bug