A Chaotic Time Series Forecasting Model Based on Parameters Simultaneous Optimization Algorithm

Abstract

Phase space reconstruction and parameters optimization were two key steps in the chaotic time series forecasting model, and tradition models could not get the optimal forecasting results without considering the relationship between the two steps. In order to improve forecasting accuracy of the chaotic time series, a chaotic time series forecasting model based on parameters simultaneous optimization was proposed. Firstly, the parameters were analyzed and designed based on uniform design, and then parameters were optimized by self-calling least square support vector machine, finally, the optimal forecasting model of chaotic time series is established, and the simulation experiment was carried out on by two chaotic time series. The simulation results show that the proposed model had improved the forecasting accuracy and computational complexity is much lower compared with the traditional models. Parameters simultaneous optimization algorithm has provided a new way to solve the parameters optimization problem of chaotic time series forecasting model.