Exploiting Collaborations in the Immune System: The Future of Artificial Immune Systems

Hart, Emma; McEwan, Chris; Davoudani, Despina

Abstract

Despite a steady increase in the application of algorithms inspired by the natural immune system to a variety of domains over the previous decade, we argue that the field of Artificial Immune Systems has yet to achieve its full potential. We suggest that two factors contribute to this; firstly, that the metaphor has been applied to insufficiently complex domains, and secondly, that isolated mechanisms that occur in the immune system have been used naïvely and out of context. We outline the properties of domains which may benefit from an immune approach and then describe a number of immune mechanisms and perspectives that are ripe for exploration from a computational perspective. In each of these mechanisms collaboration plays a key role. The concepts are illustrated using two exemplars of practical applications of selected mechanisms from the domains of machine learning and wireless sensor networks. The article suggests that exploiting the collaborations that occur between actors and signals in the immune system will lead to a new generation of engineered systems that are fit for purpose in the same way as their biological counterparts.

Citation

Hart, E., McEwan, C., & Davoudani, D. (2009). Exploiting Collaborations in the Immune System: The Future of Artificial Immune Systems. In C. Mumford, & L. Jain (Eds.), Intelligent Systems Reference Library; Computational Intelligence (527-558). Springer-Verlag. https://doi.org/10.1007/978-3-642-01799-5_16

• Publication Date: 2009
• Deposit Date: Mar 31, 2010
• Electronic ISSN: 1868-4408
• Peer Reviewed: Peer Reviewed
• Volume: 1
• Pages: 527-558
• Book Title: Intelligent Systems Reference Library; Computational Intelligence
• ISBN: 9783642017988; 9783642017995
• DOI: https://doi.org/10.1007/978-3-642-01799-5_16
• Keywords: immune system; AIS; engineered systems; complex domains
• Public URL: http://researchrepository.napier.ac.uk/id/eprint/3680
• Publisher URL: http://dx.doi.org/10.1007/978-3-642-01799-5_16