the ability of antibodies to distinguish between Self and Nonself.
Published in Chapter:
Artificial Immune Systems for Anomaly Detection
Eduard Plett (Kansas State University at Salina, USA), Sanjoy Das (Kansas State University, USA), Dapeng Li (Kansas State University, USA), and Bijaya K. Panigrahi (Indian Institute of Technology, India)
Copyright: © 2010
|Pages: 19
DOI: 10.4018/978-1-60566-766-9.ch005
Abstract
This chapter introduces anomaly detection algorithms analogous to methods employed by the vertebrate immune system, with an emphasis on engineering applications. The basic negative selection approach, as well as its major extensions, is introduced. The chapter next proposes a novel scheme to classify all algorithmic extensions of negative selection into three basic classes: self-organization, evolution, and proliferation. In order to illustrate the effectiveness of negative selection based algorithms, one recent algorithm, the proliferating V-detectors method, is taken up for further study. It is applied to a real world anomaly detection problem in engineering, that of automatic testing of bearing machines. As anomaly detection can be considered as a binary classification problem, in order to further show the usefulness of negative selection, this algorithm is then modified to address a four-category problem, namely the classification of power signals based on the type of disturbance.