Non-stationary and stationary prediction of financial time series using dynamic ridge polynomial neural network

Abstract

This research focuses on using various higher order neural networks (HONNs) to predict the upcoming trends of financial signals. Two HONNs models: the Pi-Sigma neural network and the ridge polynomial neural network were used. Furthermore, a novel HONN architecture which combines the properties of both higher order and recurrent neural network was constructed, and is called dynamic ridge polynomial neural network (DRPNN). Extensive simulations for the prediction of one and five steps ahead of financial signals were performed. Simulation results indicate that DRPNN in most cases demonstrated advantages in capturing chaotic movement in the signals with an improvement in the profit return and rapid convergence over other network models.