Bio-inspired optimization consists of drawing inspiration from the behavior and internal functioning of physical, biological and social systems to design enhanced optimization algorithms. Our aim in this paper is to enhance the performance of particle swarm optimization (PSO), when optimizing the feature selection (FS) problem, by combining two bio-inspired approaches which are chunking and cooperative learning. The experimental results show that our hybrid wrapper-filter approach gives competitive results in terms of the predictive accuracy on the training set.