摘要
孤立点检测是数据挖掘研究中的一项重要内容,其目标是发现数据集中行为异常的数据对象.本文在局部稀疏系数算法的基础上提出了基于局部最大距离的局部孤立点检测算法,该算法提出检测孤立点只需计算它的最近邻居对象的最大距离.实验结果表明,该算法发现局部孤立点是高效的.
Outlier detection, which aim is to find the abnormal objects in data set, is one of major parts of data mining.In this paper,based on algorithms for mining local sparsity coefficient-based outliers,the authors present algorithms for mining local maximum distance-based outliers Outliers are identified only by their nearest neighbor's local maximum distances in this algorithms Experiments have shown that this approach can find the local outliers effectively
出处
《河南教育学院学报（自然科学版）》
2005年第1期55-58,共4页
Journal of Henan Institute of Education(Natural Science Edition)
关键词
孤立点检测
算法
k-距离
outlier detection
algorithms
k-distance
