Please use this identifier to cite or link to this item: https://hdl.handle.net/10216/101340
Author(s): mário joão gonçalves antunes
Manuel E Correia
Title: Self Tolerance by Tuning T-Cell Activation: An Artificial Immune System for Anomaly Detection.
Issue Date: 2010
Abstract: We are currently witnessing the widespread usage of several bio-inspired metaphors as a successful source of inspiration for evolutionary and biological based learning systems. Besides the already classical examples of neural networks and genetic algorithms, the intrinsic features employed by the vertebrate immune system to protect the body constitute nowadays a major source of inspiration for the development of effective Artificial Immune Systems (AIS). These are currently employed, for example, in tasks related with network intrusion detection, spam email classification and removal, among others. The AIS constitute an emerging and verypromising area of research that historically have been falling within two main theoretical immunological schools of thought: those based on Negative selection (NS) or those based on Danger theory (DT). Despite their inherent strengths and well known promising results, NS or DT based AIS do not scale well and are known to have major difficulties on dealing with gradual dynamic environmental changes of what should be recognized as normal throughout time. In this paper we propose and describe the development of an AIS framework for anomaly detection based on the Grossman's Tunable Activation Thresholds (TAT) theory for the behaviour of T-cells. The overall framework has been tested with artificially generated stochastic data sets based on real world phenomena and the results thus obtained have been compared with the well known Support Vector Machine (SVM) classifier, thus demonstrating TAT's performance and competitiveness for anomaly detection.
Subject: Ciências da computação e da informação
Computer and information sciences
Scientific areas: Ciências exactas e naturais::Ciências da computação e da informação
Natural sciences::Computer and information sciences
URI: https://repositorio-aberto.up.pt/handle/10216/101340
Source: BIONETICS
Document Type: Artigo em Livro de Atas de Conferência Internacional
Rights: restrictedAccess
Appears in Collections:FCUP - Artigo em Livro de Atas de Conferência Internacional

Files in This Item:
File Description SizeFormat 
48787.pdf
  Restricted Access
Artigo222.69 kBAdobe PDF    Request a copy from the Author(s)


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.