摘要
对用动力系统方法进行非线性时间序列趋势项预测的问题进行了探讨.引入Copula函数构造的误差衡量模型,对几种时间序列趋势项的提取方法进行了比较,在比较准则的构造上有所创新.进而从时间序列中提取的趋势项为数据,通过动力系统方法对证券市场指数的预测进行了实证.
Prediction of trends for time series is investigated by the dynamical systems methods.Several methods of extracting the trends of time series are compared by the aid of the error measurement model derived from Copula functions.The constructing method of the comparison criterion is innovated.Using the trends of time series data,the prediction of the stock market index through the dynamical system demonstrates the validity of this method.
出处
《大连理工大学学报》
EI
CAS
CSCD
北大核心
2010年第5期831-837,共7页
Journal of Dalian University of Technology
基金
大连理工大学"数学+X"资助项目(DUT10JS06)
