Time Series Simulation using Ultra High Frequency Cosine and Cosine Higher Order Neural Networks

M. Zhang (USA)

Keywords

Analysis, Ultra High Frequency, and Artificial HONNs

Abstract

This paper presents new nonlinear models of Ultra high frequency Cosine and Cosine Higher Order Neural Networks (UCCHONN) for time series data analysis. This study also develops a new learning algorithm for UCCHONN. Based on the UCCHONN models, a time series analysis systemUCCHONN Simulator, is developed. Results show that UCCHONN models are 4.4457% to 9.0276% better than Polynomial Higher Order Neural Network (PHONN) and Trigonometric Higher Order Neural Network (THONN) models. For time series data simulation, the error rate that UCCHONN can reach is 0.0000%.

Important Links:



Go Back