Solving the Bose–Hubbard Model with Machine Learning

2017/07/31 by Hiroki Saito · 2 citations
Materials Science · Physics and Astronomy · #Artificial neural network #Cold Atom Physics and Bose-Einstein Condensates #Deep neural networks #Feedforward neural network #Ground state #Machine Learning in Materials Science #Quantum #Quantum computer #Quantum many-body systems #Quantum state #State (computer science) #cond-mat.dis-nn #cond-mat.quant-gas

paper · pdf · doi:10.7566/jpsj.86.093001

published as J. Phys. Soc. Jpn. 86, 093001 (2017) · 4 pages, 4 figures

arxiv created 2017/07/31 · openalex publication_date 2017/07/31 · arxiv updated 2017/08/01 · openalex created_date 2017/08/08 · openalex updated_date 2026/08/05

Abstract

Motivated by the recent successful application of artificial neural networks to quantum many-body problems [G. Carleo and M. Troyer, Science \bf 355, 602 (2017)], a method to calculate the ground state of the Bose-Hubbard model using a feedforward neural network is proposed. The results are in good agreement with those obtained by exact diagonalization and the Gutzwiller approximation. The method of neural-network quantum states is promising for solving quantum many-body problems of ultracold atoms in optical lattices.

Citations

• Mastering the game of Go with deep neural networks and tree search
• Solving the Quantum Many-Body Problem with Artificial Neural Networks

Cited by

• Chebyshev expansion of spectral functions using restricted Boltzmann machines
• Machine-learning quantum mechanics: Solving quantum mechanics problems using radial basis function networks