Accueil > Actualités > Séminaires > Séminaire de Davide Faranda au LMD-ENS

Séminaire

Titre : Perils of using Recurrent Neural Network to learn the Dynamics of Geophysical Flows
Nom du conférencier : Davide Faranda
Son affiliation : LSCE-IPSL
Laboratoire organisateur : LMD
Date et heure : 15-01-2020 11h00
Lieu : Salle Froidevaux - 314 - 24 rue Lhomond - 3ème étage (bouton ascenseur 6)
Résumé :

Recent advances in statistical learning have opened the possibility to forecast the behavior of chaotic systems using recurrent neural network. In this letter we investigate the applicability of this framework to geophysical flows, known to be intermittent and turbulent. We show that both turbulence and intermittency introduce severe limitations on the applicability of recurrent neural network both for short term forecasts as well as for the reconstruction of the underlying attractor. We suggest that possible strategies to overcome such limitations should be based on separating the smooth large scale dynamics, from the intemittent/turbulent features.



Authors: D. Faranda, A. Hamid, G. Carella, C.G. Ngoungue Langue, F.M.E. Pons, M. Vrac, S. Thao, M. Rabarivola, V. Gautard, P. Yiou

Contact :

helene.rouby@ens.fr