Dipartimento di Scienze Statistiche

    Lunedì 14 ottobre   Aula 34 DSS   h 11,00-12,00     Prof. Andriy Olenko (Department of Mathematical and Physical Sciences, La Trobe University, Melbourne, Australia)   This presentation discusses the semiparametric estimation of cyclic seasonal long- memory functional time series. We consider a semiparametric model of the Gegenbauer type, proposing estimates for both the singularity location and long-memory parameters, utilizing general filter transformations.  These transformations include wavelet transformations as a specific case. It is demonstrated that the proposed estimators almost surely converge to the true parameter values. Also, the asymptotic normality of these estimators is established. The solutions to the estimation equations are examined, and adjusted statistics are proposed. The results are also applicable to discretely sampled time series. Numerical studies are provided to illustrate the theoretical conclusions. The talk is based on joint results with Profs. A. Ayache and M. Fradon (Université de Lille, France) and Dr. R. Nanayakkara (La Trobe University, Australia).   e-mail: luisa.beghin@uniroma1.it