Dynamic Clone Population Based Artificial Immune System

Clonal Selection Theory indicates that only the highest affinity cells are selected to proliferate by cloning. Unlike the constant clone population in CIONALG, we propose a novel algorithm based on dynamic clone population. The novel algorithm dynamically updates the clone population by replacing low affinity cells with high affinity ones. Simulation results show that the novel algorithm has effective learning abilities and better performance when compared to CLONALG for noisy pattern recognition.