Seminars IFIC

IFIC Topical Seminar: Hierarchical autoregressive approach to two-dimensional statistical systems

by Piotr Korcyl (Jagiellonian U.)

Europe/Madrid
1001-Primera-1-1-1 - Paterna. Seminario (Universe)

1001-Primera-1-1-1 - Paterna. Seminario

Universe

60
Description

The incorporation of machine learning techniques into Monte Carlo algorithms may provide ways to improve the efficiency of such simulations. Autoregressive neural networks represent one of the approaches devised for systems with discrete degrees of freedom. I will describe our proposal to improve the scaling with the volume and will present numerical results obtained using the 12-state Potts model as an application to a system with a first-order phase transition.
 

IFIC seminar organizers