ADAPTIVE ESTIMATES FOR PARFIMA MODELS
Amimour Amine  1, 2, @  
1 : National Institute for Research in Education  -  Website
Oued El Rouman, Alger, Algeria. -  Algeria
2 : Université de Bejaïa  -  Website

In this paper, we consider periodically time-varying autoregressive fractionally integrated moving average processes (PARFIMA). We construct adaptive estimates for a periodic long memory parameter using the Local Asymptotic Normality (LAN) property. The performance of the achieved adaptive estimator is shown by a simulation study.

Subject : : Oral
Topics : Probability ans statistics
Keywords : LAN property ; Long memory process ; Adaptive estimate ; Locally Asymptotically Minimax ; Periodic ARFIMA models ; Simulation.
Online user: 58 Privacy
Loading...