IRAP > Séminaires > Calendrier des séminaires > Deep Learning in Solar Physics

Deep Learning in Solar Physics

Séminaire le 16 nov 2017 de 11h00 à 12h00

Intervenant : Andrés Asensio Ramos

IAC, Espagne

Salle de conférence Salle de conférence

Deep learning has emerged as a very powerful set of techniques to extract relevant information from observations, sometimes showing much better results that other set of finely tuned algorithms. In this contribution I present our efforts in applying deep learning and other techniques based on sparsity regularization methods to several problems in Solar Physics, from the estimation of horizontal velocities in the solar surface to fast image reconstruction.

Afficher le pied de page