Research on outlier detection for high dimensional data stream

DOI

10.2991/aiea-16.2016.70How to use a DOI?

Keywords

High dimensional; data stream; outlier detection.

Abstract

The development of the Internet of things has put forward new requirements to the data processing capacity, and outlier detection has found an increasingly wide utilization in the field of data mining. The accuracy of the outlier detection algorithm based on Euclidean distance in the high dimensional data detection cannot be guaranteed, what is worse, the processing time is too long. This paper constructs the small data sets of the best set of data grid and recently data grid, in order to calculate the abnormal degree of the newest data point by measuring angle variance of the high dimensional data stream; as data stream capture, the best data grid and data grid updated incently, whose aim is to solve the concept transferring of big data flow. The experimental results show that compared with the ABOD algorithm and the classical algorithm, this algorithm is more suitable for the outlier detection of the high dimensional data stream in the Internet of things.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Liping Yu
AU  - Yunfei Li
AU  - Juncheng Jia
PY  - 2016/11
DA  - 2016/11
TI  - Research on outlier detection for high dimensional data stream
BT  - Proceedings of the 2016 International Conference on Artificial Intelligence and Engineering Applications
PB  - Atlantis Press
SP  - 395
EP  - 398
SN  - 2352-538X
UR  - https://doi.org/10.2991/aiea-16.2016.70
DO  - 10.2991/aiea-16.2016.70
ID  - Yu2016/11
ER  -