The review of outlier mining based on granular computing

Outlier mining is an important part in data mining. In this paper, through the research of causes and detection methods of outlier, combined with the perspective of granular computing, we firstly present a general guiding principle of outlier mining that is granulation viewpoint, which shows that ch...

Full description

Saved in:
Bibliographic Details
Main Authors: Shuang Liu, Jiyi Wang, Guolin Xing
Format: Conference Proceeding
Language:English
Subjects:
Algorithm design and analysis
Data mining
granular computing
granularity
granulation
Humans
outlier detection method
outlier mining
Partitioning algorithms
Presses
Problem-solving
Technological innovation
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Outlier mining is an important part in data mining. In this paper, through the research of causes and detection methods of outlier, combined with the perspective of granular computing, we firstly present a general guiding principle of outlier mining that is granulation viewpoint, which shows that choosing reasonable granularity before granulation plays a very crucial role in outlier mining. Then we show a unified process frame diagram based on granular computing for outlier detection. Finally, a new algorithm based on granulation viewpoint for outlier detection is given. We argue that this paper can provide the practical reference value for the selection, improvement and the innovation of outlier detection method. And outlier mining based on granular computing will offer a kind of new thinking of research and analytic method for the future research topics and the challenges of the outlier mining.
DOI:10.1109/GRC.2008.4664651