首页 | 本学科首页   官方微博 | 高级检索  
     检索      


Autoregressiv model identification for multivariate time series
Abstract:A procedure is developed for the identification of autoregressive models for stationary invertible multivariate Gaussian time series. Model selection is based on either the AIC information criterion or on a statistic called CVR, cross-validatory residual sum of squares. An example is given to show that the forecasts generated by these models compare favorably with those generated by other common time series modeling techniques.
Keywords:Multivariate time series  autoregressiv models  model identification  cross-validation
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号