摘要
本文在平稳和非平稳情形,构造了两指标时间序列的一类预报方法。在平稳条件下构造了两指标 AR 过程的最佳线性预报。在非平稳情形,利用单指标时间序列的新息预报思想,首先将非平稳过程的自回归系数看作是一个新的两指标多维平稳过程加以预报,然后再对非平稳过程本身做最佳线性预报。
Under stationary and unstationary conditions,a class of forecasting methods oftwo-parameter time series is put forward.Under the stationary condition,the best linearforecast of two-parameter AR process is constructed,and under the unstationary condi-tion,the thought of new-information forecasting of single—parameter time series is used,primarily,to treat autoregressive coefficients as a two-parameter-stationary multiple se-quence to be forecasted,and secondly,to construct a best linear forecast of theunstationary process itself.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
1991年第2期71-77,共7页
Journal of Xidian University
关键词
AR过程
时间系列
线性预报
two-parameter stationary procsess
two-parameter linear-stationary process
two-parameter AR process
best linear forecast
