Financial prediction using higher order trigonometric polynomial neural network group model

Abstract

A higher order trigonometric polynomial neural network group model (HTG) which can be used for financial prediction is discussed in this paper. HTG is written in C, incorporates a user-friendly graphical user interface, and runs under XWindows on a Sun workstation. The experimental results show that HTG is able to handle higher frequency, higher order nonlinear and discontinuous data. The accuracy of HTG is around 5% to 10% better than conventional trigonometric polynomial neural network models.