Studies on the Implications of Shape-Space Models for Idiotypic Networks

Abstract

Despite the flood of practical work in the AIS area over the past few years, unlike in other fields derived from biological paradigms, there has yet to emerge a concise view of what an artificial immune system actually is, although most interpretations of the metaphor possess a set of common elements. In this conceptual paper, we argue that one of the components of an AIS algorithm that most distinguishes it from other paradigms is the matching aspect (although memory, as argued by [9], is clearly also a defining feature of AIS algorithms). However, for the most part, the implications of choosing a suitable matching rule and the likely effects on the algorithm from an engineering point of view have not been well studied. We present some results obtained using freely available software that simulates an idiotypic network with various models of shape-space, in the hope of illustrating the importance of selecting a suitable shape-space and matching rule in practical implementions of AISs. We raise some interesting questions and encourage others to experiment with the tool also in order to better understand the performance of their algorithms and to be able to design better ones.