Re-editing and Censoring of Detectors in Negative Selection Algorithm

DOI

10.2991/ijcis.2009.2.3.11How to use a DOI?

Abstract

The Negative Selection Algorithm (NSA) is a kind of novelty detection method inspired by the biological self/nonself discrimination principles. In this paper, we propose two new schemes for the detectors re-editing and censoring in the NSA. The detectors that fail to pass the negative selection phase are re-edited and updated to become qualified using the Differential Evolution (DE) method. In the detectors censoring, the qualification of all the detectors is evaluated, and only those appropriate ones are retained. Prior knowledge of the anomalous data is utilized to discriminate the detectors so that their anomaly detection performances can be improved. The effectiveness of our detectors re-editing and censoring approaches is examined with both artificial signals and a practical bearings fault detection problem.

Copyright

© 2009, 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  - JOUR
AU  - X.Z. Gao
AU  - S. J. Ovaska
AU  - X. Wang
PY  - 2009
DA  - 2009/10/01
TI  - Re-editing and Censoring of Detectors in Negative Selection Algorithm
JO  - International Journal of Computational Intelligence Systems
SP  - 298
EP  - 311
VL  - 2
IS  - 3
SN  - 1875-6883
UR  - https://doi.org/10.2991/ijcis.2009.2.3.11
DO  - 10.2991/ijcis.2009.2.3.11
ID  - Gao2009
ER  -