A New Guidance for Optimizing the Artificial Neural Networks Predictive Ability in the RMB Ezchange Rate Series Forecasting
Accuracy prediction of the RMB exchange rate is one of the most primary issues to avoid risk for China during the recent global financial tsunami. ANNs have already been a general concerned tool in exchange rate forecasting applications in recent years. However, not only for its own drawbacks, but also because that the RMB exchange rate behavior has become more complex since the Clunese exchange rate mechanism reform, which makes it much more difficult for a single ANN model to learn the underlying pattern well and offer accuracy prediction. Motivated by hybrid methodologies, this research put forward a new guidance for optimizing the ANNs predictive ability, which is to combine the ANN model with EMd technique and AC approach to study on CNY/USd exchange rate volatility series. The empirical results show that the proposed method can significantly optimize the ANN model, which is more capable than the single MLP and simple random walk model, especially in RMB exchange rate volatility forecasting.
ANNs EMD technique AC approach RMB ezchange rate forecasting
Chi Xie Bo Sun Juan Zhang
Center of Finance and Investment Management, Hunan University, Hunan 410082, P.R.China College of Business Management, Hunan University, Hunan 410082, P.R. China
国际会议
The First World Congress on Global Optimization in Engineering & Science(第一届工程与科学全局优化国际会议 WCGO2009)
长沙
英文
828-837
2009-06-01（万方平台首次上网日期，不代表论文的发表时间）