摘要
本文尝试用人工神经网络方法(Artificial Neural Network,简称ANN)预测USD/DEM汇率.研究结果表明,人工神经网络算法用于汇率预测是有效的(与随机游走模型相比).模型中的宏观经济数据采用取对数之差的处理方法,大大扩大了模型的应用范围.
In this paper,Artificial Neural Network (ANN) is adopted to forecast the Changes of Exchange Rate of USD/DEM. From this research we can draw the conclusion that ANN is much more efficient .through that comparison is made between the forecasting performance of ANN and that of RW model. In addition, the pre-processing method of date adopted in this paper takes model to a more wide extent.
出处
《系统工程》
CSCD
1999年第1期18-24,共7页
Systems Engineering
基金
国家自然科学基金资助项目
关键词
人工神经网络
中长期汇率
汇率预测
外汇市场
Artificial Neural Network (ANN ), random walk model (RW) , mid-long term exchange rate forecasting, the forecastanility of exchange rate,backpropagation