Open Access System for Information Sharing

Login Library

 

Article
Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

Accelerated continuous time quantum Monte Carlo method with machine learning

Title
Accelerated continuous time quantum Monte Carlo method with machine learning
Authors
SONG, TAEGEUNLee, Hunpyo
Date Issued
Jul-2019
Publisher
AMER PHYSICAL SOC
Abstract
An acceleration of continuous time quantum Monte Carlo (CTQMC) methods is a potentially interesting branch of work as they are matchless as impurity solvers of a dynamical mean field theory (DMFT) approach for the description of strongly correlated systems. The inversion of the k x k matrix with k(2) operations given by the diagram expansion order k in the CTQMC fast update and the multiplication of the k x k matrix, and the noninteracting properties with k x omega(m-1)(nmax) operations to measure the in-point correlators, are computationally time consuming. Here we propose the CTQMC method in combination with a machine learning technique, which eliminates the k x omega(nmax) and k x omega(3)(nmax) operations for the two-point impurity Green's functions G(sigma) (i omega(n)) and four-point vertices chi(sigma), ((sigma) over bar) (i omega(n1), i omega(n2), i omega(n3), i omega(n4)), respectively. This method not only predicts the accurate physical properties at low temperature, but also dramatically decreases the computational times of chi(sigma), ((sigma) over bar) (i omega(n1), i omega(n2), i omega(n3), i omega(n4)) for the nonlocal extension of DMFT approximation.
Keywords
Computation theory; Continuous time systems; Correlators; Machine learning; Mean field theory; Molecular physics; Temperature; Computational time; Continuous-time; Diagram expansion; Dynamical mean-field theory; Low temperatures; Machine learning techniques; Noninteracting; Strongly correlated systems; Monte Carlo methods
URI
http://oasis.postech.ac.kr/handle/2014.oak/99555
ISSN
2469-9950
Article Type
Article
Citation
PHYSICAL REVIEW B, vol. 100, no. 4, 2019-07
Files in This Item:
There are no files associated with this item.

qr_code

  • mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Views & Downloads

Browse