Abstract:
Outlier mining, aim of which is to discover the abnormal data objects in the data set, is an important research aspect in the data mining. The traditional outlier mining algorithms are based on the characteristics of item set, and unsuitable for the application in the multi-objective decision and synthetical evaluation. This paper presents an outlier mining algorithm based on the grey relation analysis. This algorithm is able to mine outliers by the number of the synthetical evaluation and avoid choosing the user-specified threshold. Experimental results show that this algorithm can detect all outliers efficiently in data set, and the mined outliers are coincident and objective.
Key words:
data mining,
outlier detection,
grey relation analysis,
relation coefficient
摘要: 孤立点挖掘是数据挖掘的重要研究方向之一,其目标是发现数据集中不具备数据一般特性的数据对象。传统孤立点挖掘算法通常基于项集属性,不适用于多目标决策和综合评价。该文提出一种基于灰色关联分析的孤立点检测算法OMGRA,通过总评价判断数挖掘孤立点集,避免人工确定阈值。实例分析表明,该算法能有效检测数据集中的孤立点,挖掘出的孤立点符合实际情况。
关键词:
数据挖掘,
孤立点检测,
灰色关联分析,
关联系数
CLC Number:
LI Yun; YUAN Yun-hao; CHEN Ling. Outlier Mining Algorithm Based on Grey Relation Analysis[J]. Computer Engineering, 2008, 34(19): 44-46.
李 云;袁运浩;陈 崚. 基于灰色关联分析的孤立点挖掘算法[J]. 计算机工程, 2008, 34(19): 44-46.