摘要
对于含有异常值的动态数据如何进行参数估计及建模,本文提出了二种方法:时间序列的稳健估计和时间序列的稳健化处理(异常值的识别、检测与剔除)。
Motivated by the presence of outliers in time series,robust analysis for time series has been developed rapidly in recent years,and is becoming more and more an attracting field.There outliers could lead analysis results such as building model,parameters estimate,testing,etc to a wrong way.In the paper,we discuss robust estimate for time series and robustization treatment of time series including discrimination, testing and picking out of outliers from series.
出处
《长沙铁道学院学报》
CSCD
1990年第1期1-13,共13页
Journal of Changsha Railway University
