A REVIEW OF CHAOS THEORY PRACTICED IN NEURAL NETWORK

Abstract

In this paper, a review of chaos theory, which has been practiced in neural network since 2000, has been presented. A short introduction for chaos theory in the field of artificial neural network has been presented to clarify connection between chaos and different neural network architectures. Phase space construction that covers the time delay, embedding dimension, Lyapunov exponent and fractal dimension has been described in the scope of the introduction of chaos theory. Moreover, neural networks, which have been mostly practiced with chaos theory, have been stated. This study aims to clarify the integrated application between chaos theory and different neural network architectures. This paper would guide future research directions in distinct fields and present the methods to be exerted in neural network.