Many Body Seminar: Machine Learning techniques to speed up Monte Carlo computations

Patrick Emonts (MPQ)
Monte Carlo sampling is commonly used in machine learning, for example to sample restricted Boltzmann machines.

November 11, 2021

Patrick Emonts (MPQ)
Group seminar (hybrid format: online/seminar room B2.46)
Thu 11 November 2021, 11:00 am (MEZ)

*Abstract:*

Monte Carlo sampling is commonly used in machine learning, for example to sample restricted Boltzmann machines.
But can we turn the idea around and use machine learning techniques to speed up Monte Carlo simulations?
The talk will be in a Journal Club style format.
We will have a look at two recent papers (10.1103/PhysRevB.102.195148, 10.1103/PhysRevE.100.043301) and see how they use restricted Boltzmann machines to speed up Monte Carlo simulations.

We meet at 11h in the Theory seminar room starting with an informal
discussion, the talk starts at 11:20. Please don't forget to do a corona self test before.