Correlations involved in a bio-inspired classification technique

Abstract

An improved unsupervised bio-inspired clustering model is introduced. The main goal is to involve a correlation between properties of objects and some bio-inspired factors. The statistical classification biological model is based on the chemical recognition system of ants. Ants are able to create groups discriminating between nest-mates and intruders based on similar odor. Comparative analysis are performed on real data sets. © 2011 Springer-Verlag Berlin Heidelberg.