An Improved Dynamic Bayesian for Exchange Rate Forecasting

Abstract: A feasibility study of using of Dynamic Bayesian Networks in combination with ARMA modeling in exchange rate prediction is presented. A new algorithm (ARMA-DBN) is constructed and applied to the exchange rate forecast of RMB. Results show that the improved dynamic Bayesian forecast algorithm has better performance than the standard ARMA model.
Keywords: Dynamic Bayesian Network, ARMA Model, Exchange Rate Forecasting
Author: Mingjuan Xu, Zhengyu Liu
Journal Code: jptkomputergg160061

Artikel Terkait :