Date of Submission

Spring 2018

Academic Programs and Concentrations

Physics; Computer Science

Project Advisor 1

Paul Cadden-Zimansky

Project Advisor 2

Sven Anderson

Abstract/Artist's Statement

One of the fundamental problems in analytically approaching the quantum many-body problem is that the amount of information needed to describe a quantum state. As the number of particles in a system grows, the amount of information needed for a full description of the system increases exponentially. A great deal of work then has gone into finding efficient approximate representations of these systems. Among the most popular techniques are Tensor Networks and Quantum Monte Carlo methods. However, one new method with a number of promising theoretical guarantees is the Neural Quantum State. This method is an adaptation of the Restricted Boltzmann machine(RBM). Unlike the traditional RBM, Neural Quantum States act as a feedforward network, calculating a single complex value of the wave function for every spin configuration. We examine this method, and compare its performance to a feedforward network for a similar problem. Another recent application includes the use of neural networks for detecting phase transitions. We examine the claims made about this technique and propose a new method for solving this problem. We report results for both experiments.

Open Access Agreement

Open Access

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

This work is protected by a Creative Commons license. Any use not permitted under that license is prohibited.