Oscillation Patterns of A Complex Exponential Neural Network

Abstract

The paper presents the design and evaluation of a complex exponential neural network model. The development of the model is inspired by the exponential form of general solutions to nonlinear differential equations that describe dynamical systems. The research goal is to develop a mathematical representation for neural oscillation and reduce the amount of computation in the neural network to improve computational efficiency. In particular, the weighted sum of two complex exponential neurons is evaluated to demonstrate that the difference of oscillation frequencies between the two neurons is the dominant parameter that determines the oscillation patterns of the neural network.