A MapReduced-Based and Cell-Based Outlier Detection Algorithm

Outlier detection is a very important type of data mining,which is extensively used in application areas.The traditional cell-based outlier detection algorithm not only takes a large amount of time in processing massive data,but also uses lots of machine resources,which results in the imbalance of t...

Full description

Saved in:
Bibliographic Details
Published in:Wuhan University journal of natural sciences 2014-06, Vol.19 (3), p.199-205
Main Authors: Zhu, Sunjing, Li, Jing, Huang, Jilin, Luo, Simin, Peng, Weiping
Format: Article
Language:English
Subjects:
Algorithms
Biomedical and Life Sciences
Computation
Computer Science
Data mining
Life Sciences
Materials Science
Parallel processing
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Outlier detection is a very important type of data mining,which is extensively used in application areas.The traditional cell-based outlier detection algorithm not only takes a large amount of time in processing massive data,but also uses lots of machine resources,which results in the imbalance of the machine load.This paper presents an algorithm of the MapReduce-based and cell-based outlier detection,combined with the single-layer perceptron,which achieves the parallelization of outlier detection.These experiments show that this improved algorithm is able to effectively improve the efficiency of the outlier detection as well as the accuracy.
ISSN:1007-1202
1993-4998
DOI:10.1007/s11859-014-1002-8