A Goodness-of-Fit Test in Robust Time Series Modelling
针对存在异常值的时间序列模型，将经典Portmanteau检验推广到稳健估计量，蒙特卡洛模拟表明新方法比经典统计量更稳健，并得到了Newbold(1980)结果的稳健版本。
The problem of testing the adequacy of a time series model in the presence of outliers is considered. The classical portmanteau statistic is generalized to an important class of robust estimators. Some Monte Carlo results suggest that the proposed generalization possesses good robustness properties over the classical statistic. A robustified version of a result of Newbold (1980) is also obtained.