Which Neural Network Is Appropriate for CNY/USD Exchange Rate Series Forecast?

摘要：

Artificial neural networks (ANNs) have received more and more attention in exchange rate series forecasting in recent years.Considering the network models perform differently for different exchange rate series,in this paper,the in- or out-of-sample predicting capability of the two typical networks with different lag parameters determined by the AC approach are compared on both CNY/USD exchange rate level and volatility series.The empirical results show that the recurrent network is a more appropriate model than the feed-forward network especially in RMB exchange rate series forecasting.

关键词： ANNs AC approach exchange rate series forecasting

作者: Chi XIE Bai SUN Juan ZHANG

作者单位: College of Business Management,Hunan University,Changsha 410082,China Center of Finance and Investme College of Business Management,Hunan University,Changsha 410082,China

会议类型: 国际会议

会议名称: The 2008 International Conference on Business Intelligence and Financial Engineering(BIFE 2008)(商业智能和金融工程国际会议)

会议地点: 长沙

会议语种:英文

页码: 114-122

在线出版日期: 2008-10-28（万方平台首次上网日期，不代表论文的发表时间）