Robust Multiple Time Series Modelling
提出了一种基于稳健化残差自协方差的多元时间序列估计方法，推导了估计量的渐近分布，并给出了可用于模型诊断的稳健多元Portmanteau统计量，以貂-麝鼠数据为例说明。
A robust estimation procedure for multiple time series is proposed based on robustifying the residual autocovariances in the estimating equation. The asymptotic distribution of these estimators is derived. A robustified multivariate portmanteau statistic is also obtained which can be useful in model diagnostic checking. An illustrative example based on the mink-muskrat data is presented.