摘要
提出了两种基于广义指数预报因子模型的石油价格预测方法.该方法使用拟合期内的样本,在不同准则下选取有限个不同参数的EWMA的线性组合,作为油价的预报因子.实证分析显示,该方法的预报结果比文献中已有的结果要准确.
This paper proposes two methods for crude oil price forecasting based on generalized exponential predictors.These methods select a finite number of EWMAs of in-sample data with different parameters, then make linear combination of these EWMAs under different criteria,as the predictor of crude oil price. The experiment results show that GEP outperforms other methods proposed in previous papers.
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2010年第8期1389-1395,共7页
Systems Engineering-Theory & Practice
关键词
指数加权估计
石油价格
指数加权滑动平均
变量筛选
时间序列预报
exponentially weighted estimating
crude oil price
exponentially weighted moving-average
variable select
time series forecasting
