摘要
提出了一种基于空间相邻关系的点对象离群检测算法SAOD(Space Adjacent Relations Based GML Point Outlier Detec-tion Algorithm).利用空间相邻关系作为空间点对象的相似度度量准则,得到相似度矩阵,从而挖掘GML中的离群点对象.实验结果表明,SAOD算法能有效地检测GML中的离群点对象并且具有较高的效率.
At present algorithm for GML outlier detection has seldom been researched. Algorithm SAOD for GML point outlier detection based on space adjacent relations is proposed in this paper. In this algorithm, the space adjacent relations between spatial points are considered as the similarity measurement and similarity matrix is computed. The expected outliers can be obtained from the matrix. The results of experiments show that algorithm SAOD is effective and efficient.
出处
《南京师范大学学报(工程技术版)》
CAS
2009年第1期61-63,共3页
Journal of Nanjing Normal University(Engineering and Technology Edition)
基金
国家自然科学基金(40771163
40871176)资助项目
关键词
离群点检测
空间相邻
GML数据挖掘
outlier detection, space adjacent relations, GML data mining