A Swarm-based Evolutionary Morphological Approach for Binary Classification Problems

Abstract

-In this work we propose a swarm-based evolutionary morphological approach to deal with binary classification problems. It consists of a hybrid neuron based on principles of mathematical morphology and lattice theory, referred to as dilation-erosion-linear perceptron (DELP). We also present a swarm-based evolutionary learning process, called DELP(PSO), using a particle swarm optimizer (PSO) to design the DELP model, due to some drawbacks from gradient estimation of morphological operators in the classical learning process of the DELP. Besides, we conduct an experimental analysis using two relevant binary classification problems and the obtained results are discussed and compared with those obtained by established techniques in the literature.