Artificial Immune Systems: A Novel Paradigm to build architectures

Omar U. Flórez Choque (1)

e-mail: omarflores19@gmail.com

(1) National University of San Agustín, Computer Science department.
Arequipa, Perú

Abstract

Actually nature-inspired solutions are broadly used, such as biological and social systems. Artificial Neural Networks (ANN), genetic algorithms (GAs) and swarm intelligence are good examples of this. Despite some isolated efforts, there are not generic frameworks with which to represent, design and understand biologicallymotivated systems, in spite of existing many similarities between the above mentioned techniques such as structure of representation, environment interaction function and learning and adaptation operators.

This biologically-motivated architecture uses agents and exposes the main goodness of a typical classifier system: Mechanism to generalize trough rules, an environment, energy or power of classification trough statistical and data mining technics, feedback from the environment and specificity. In these aspects resemble to Artificial Neural Networks.

PDF de este artículo

PDF de JPC2006 (incluye todos los artículos)