Modeling Nonlinear Dynamics with Neural Networks: Examples in Time Series Prediction
Eric A. Wan · 1993
A neural network architecture is discussed which uses Finite Impulse Response (FIR) linear filters to provide dynamic interconnectivity between processing units. The network is applied to a variety of chaotic time series prediction tasks. Phase-space plots of the network dynamics are given to illustrate the reconstruction of underlying chaotic attractors. An example taken from the Santa Fe Institute Time Series Prediction Competition is also presented.