摘要
针对黄金价格时间序列的非平稳性特征,将小波分析与FAR(函数系数自回归)模型结合,并作预测分析.利用Mallat算法中的Daubechies小波变换和多项式样条估计,对1991年1月42007年12月的厅平均国际黄金价格,建立了基于小波变换的FAR模型,并对2008年1月-2008年3月的数据进行短期预测.预测误差明显减小.在国际黄金价格的预测中,基于小波变换的FAR模型优于单纯的FAR模型.
Due to the non-stability of gold prices time series, build the FAR model combined with wavelet analysis. Use Daubechies Wavelet of the Mallat algorithm and the multi-sample estimate to build the functional-coefficient autoregression model of international gold prices based on wavelet analysis. The data of monthly average international gold prices from January 1991 to December 2007 are used to estimate and predict the monthly average international gold prices from January 2008 to March 2008. The forecast error reduced obviously. The results shows that the FAR model based on the wavelet transform is superior to the FAR model.
出处
《西安工业大学学报》
CAS
2009年第1期84-88,共5页
Journal of Xi’an Technological University