2018 Beg Rohu summer school: Deep Learning and Statistical Physics  


Registration to the 2018 Beg Rohu summer school of statistical physics and condensed matter: Deep Learning and Statistical Physics is open. The deadline for application is the 25th of March.

The school will take place from the 25th of June to the 7th of July 2018 at the French National Sailing School, on the peninsula of Quiberon in southern Brittany, France.

Lectures: G. Ben Arous: Complexity of random landscapes, L. Bottou: Training Deep Networks with Stochastic Gradient Algorithms, J. Kurchan: Rugged phase-spaces and the dynamics within them, S. Mallat: Learning Physics with Multiscale Deep Neural Networks, M. Mézard: Statistical physics of inference, H. Sompolinsky: Statistical Mechanics of Deep Neural Networks and seminars by F. Krzakala and R. Zecchina

Organizer: G. Biroli, corganizers: C. Cammarota and M. Tarzia

More detailed information can be found on the school webpage: http://ipht.fr/begrohu

A special care will be devoted to pedagogy: there will be long series of lectures on the blackboard that will go progressively from the basics to the current research edge. Furthermore, there will be sailing lessons in the afternoon.


F. David, dépêche du 12/02/2018


Retour en haut