In this paper we discuss a preliminary results on the construction of a weighted likelihood procedure for robust estimation of the unknown parameters of an autoregressive-moving average model. Two types of outliers, i.e., additive and innovations be take into account without knowing their number, position or intensity. A classification procedure based on a selection criterion is used to identified the most useful pattern rappresentation of the outliers to gain efficiency and robustness. Two examples are reported.
Robust time series estimation via weighted likelihood. Some preliminary results.
Agostinelli, Claudio
2001
Abstract
In this paper we discuss a preliminary results on the construction of a weighted likelihood procedure for robust estimation of the unknown parameters of an autoregressive-moving average model. Two types of outliers, i.e., additive and innovations be take into account without knowing their number, position or intensity. A classification procedure based on a selection criterion is used to identified the most useful pattern rappresentation of the outliers to gain efficiency and robustness. Two examples are reported.File in questo prodotto:
File | Dimensione | Formato | |
---|---|---|---|
35_pdfsam_WP_2001_parte_1.pdf
accesso aperto
Licenza:
Non specificato
Dimensione
30.22 MB
Formato
Adobe PDF
|
30.22 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.