计算机工程 ›› 2009, Vol. 35 ›› Issue (19): 94-96.doi: 10.3969/j.issn.1000-3428.2009.19.031

• 软件技术与数据库 • 上一篇    下一篇

MELOF算法的理论分析与拓展

李 健1,2,阎保平1,李 俊1   

  1. (1. 中国科学院计算机网络信息中心,北京 100190;2. 中国科学院研究生院,北京 100190)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-10-05 发布日期:2009-10-05

Theoretical Analysis and Development of MELOF Algorithm

LI Jian1,2, YAN Bao-ping1, LI Jun1   

  1. (1. Computer Network Information Center, Chinese Academy of Sciences, Beijing 100190; 2. Graduate University of Chinese Academy of Sciences, Beijing 100190)
  • Received:1900-01-01 Revised:1900-01-01 Online:2009-10-05 Published:2009-10-05

摘要: 介绍LOF算法、记忆效应以及MELOF算法,对记忆效应进行理论证明,验证MELOF算法的正确性,同时分析该算法的不足和记忆效应的一些特性。针对MELOF算法中的不足进行改进,介绍未来的研究方向,即参数自动选择和利用分而治之思想提高运行效率等。

关键词: 数据挖掘, 异常检测, 局部异常因子, 记忆效应, MELOF算法

Abstract: After introducing Local Outlier Factor(LOF) algorithm, memory effect and MELOF algorithm, this paper gives the theoretical analysis of memory effect. According to the analysis, it proves the correctness of MELOF algorithm, and finds both the shortcoming of MELOF algorithm and some features of memory effect. It improves MELOF algorithm. The research areas in the future, such as selecting proper parameter value automatically and rising the efficiency with the help of divide and conquer method are introduced.

Key words: data mining, outlier detection, Local Outlier Factor(LOF), memory effect, MELOF algorithm

中图分类号: