Ermal Rrapaj, University of Berkeley - University of Minnesota
Thursday, February 6, 2020 - 2:30pm
In recent decades, machine learning in the form of deep neural networks, has been effectively used in diverse industrial applications. In physics, deep learning has been used to analyse particle accelerator data, discover phases of matter, represent ground states of quantum many body systems, and accelerate numerical algorithms. This talk will focus on the Restricted Boltzmann Machine and how it can be used to derive exact representations of many body forces with applications in quantum Monte Carlo simulations and quantum annealing. I conclude with implications for training this type of architecture based on its connection to physical systems.