Partenaires

CNRS
Logo tutelle
Logo tutelle
Logo tutelle


Rechercher

Sur ce site

Sur le Web du CNRS


Accueil du site > Séminaires > Probabilités Statistiques et réseaux de neurones > Artificial Neural Network and Semiparametric Long-Memory ARCH

Vendredi 27 mai 2005, à 10h

Artificial Neural Network and Semiparametric Long-Memory ARCH

Gilles Teyssière (Samos, Université Paris 1)

Résumé : We propose a new class of processes, the Long-Memory Nonlinear Asymmetric GARCH (LM-NGARCH), which extend the Nonlinear Asymmetric GARCH by engle:1990 to fractional situations. We also consider the stationary semi long-memory ARCH processes, which mix hyperbolic and exponential rate of decay of the lag polynomials, and the Artificial Neural Network Long—Memory ARCH. We apply these models to the series of log of returns on the S\&P500 index. We compare the forecasting performance of these models.

Dans la même rubrique :