Self-Organization and Emergence in Neural Networks

The interest for neural networks stems from the fact that they appear as universal approximators for whatever kind of nonlinear dynamical system of arbitrary complexity. Because nonlinear systems, are studied mainly to model self-organization and emergence phenomena, there is the hope that neural networks can be used to investigate these phenomena in a simpler way, rather than resorting to nonlinear mathematics, with its unsolvable problems. In this contribution we will deal with the conditions granting for the occurring of selforganization and emergence phenomena in neural networks. We will present a number of arguments supporting the claim that emergence is strictly connected to the presence of quantum features in neural networks models. c © Electronic Journal of Theoretical Physics. All rights reserved.