空间孤立点检测

空间孤立点是指与邻居具有不连续性的空间点，或者是偏离观测值以至使人们认为是由不同的体系产生的。空间孤立点检测在交通、生态、公共安全、卫生健康、地震、海啸等领域有广泛应用。传统的根据一个非空间属性值进行孤立点判断的方法客易引起孤立点判断失误。作者在针对多个属性进行考虑的基础上，提出以空间维确定邻居关系，非空间维定义距离函数，使用Mahalanobis距离检测孤立点，研究一种新的检测空间孤立点的算法。并时时间复杂度进行分析。仿真实验说明算法可以有效地发现大规模空间数据中的孤立点。

Spatial Outlier Detection Algorithm

Spatial outliers have been informally defined as observations in a data set which appear to be inconsistent with the remainder of that set of spatial data,or which deviate so much from other observations so as to arouse suspicions that they were generated by a different mechanism. Spatial outliers detection is widely used in credit card, fraud detection, public sanitation analysis and tsunami predication and many other fields. Based on multi-attributes, the spatial attributes are used to determine neighborship and non-spatial attributes to define distance function in the paper. A novel algorithm detecting spatial outliers is proposed, which employs Mahalanobis distance to detect outliers. Computation time complexity is analyzed. The simulated experiments demonstrate that our approach can effectively identify local abnormality in large spatial data sets.