Towards efficient and effective negative selection algorithm: a convex hull representation scheme

Authors:
Mahshid Majd;Farzaneh Shoeleh;Ali Hamzeh;Sattar Hashemi
Affiliations:
Computer Science and Engineering Dept., Shiraz University, Shiraz, Iran;Computer Science and Engineering Dept., Shiraz University, Shiraz, Iran;Computer Science and Engineering Dept., Shiraz University, Shiraz, Iran;Computer Science and Engineering Dept., Shiraz University, Shiraz, Iran
Venue:
SEAL'10 Proceedings of the 8th international conference on Simulated evolution and learning
Year:
2010

Citing 8
Cited 0

Self-Nonself Discrimination in a Computer

SP '94 Proceedings of the 1994 IEEE Symposium on Security and Privacy
An Immunological Approach to Change Detection: Algorithms, Analysis and Implications

SP '96 Proceedings of the 1996 IEEE Symposium on Security and Privacy
Anomaly Detection Using Real-Valued Negative Selection

Genetic Programming and Evolvable Machines
An evolutionary algorithm to generate hyper-ellipsoid detectors for negative selection

GECCO '05 Proceedings of the 7th annual conference on Genetic and evolutionary computation
Using convex hulls to represent classifier conditions

Proceedings of the 8th annual conference on Genetic and evolutionary computation
Revisiting Negative Selection Algorithms

Evolutionary Computation
V-detector: An efficient negative selection algorithm with "probably adequate" detector coverage

Information Sciences: an International Journal
An immunity-based technique to characterize intrusions in computernetworks

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Negative Selection Algorithm (NSA) is one of several algorithms inspired by the principles of natural immune system. The algorithm received the researchers attention due to its applicability in various research areas and a number of valuable efforts are made to increase the effectiveness and efficiency of it. The heart of NSA is to somehow find rules called detectors to discriminate self and anomaly areas. Each detector in NSA defines a subspace of problem space where no self data is located. One of the major issues in NSA is detector's shape or representation of detectors which can affect the detection performance significantly. This paper for the first time proposes a new representation for detectors based on convex hull. Since convex hull is a general form of other geometric shapes, it retains the benefits of other shapes meanwhile it provides some new features like the asymmetric shape. Experimental results show a significant enhancement in the accuracy of negative selection algorithm compared to other common representation shapes.