Bayesian Optimization of Bose-Einstein Condensates
Abstract Machine Learning methods are emerging as faster and efficient alternatives to numerical simulation techniques. The field of Scientific Computing has started adopting these data-driven approaches to faithfully model physical phenomena using scattered, noisy observations from coarse-grained g...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Publishing Group
2021-03-01
|
Series: | Scientific Reports |
Online Access: | https://doi.org/10.1038/s41598-021-84336-0 |
id |
doaj-e2e87a1639044c67a46590d351cc9e6e |
---|---|
record_format |
Article |
spelling |
doaj-e2e87a1639044c67a46590d351cc9e6e2021-03-11T12:13:47ZengNature Publishing GroupScientific Reports2045-23222021-03-011111910.1038/s41598-021-84336-0Bayesian Optimization of Bose-Einstein CondensatesTamil Arasan Bakthavatchalam0Suriyadeepan Ramamoorthy1Malaikannan Sankarasubbu2Radha Ramaswamy3Vijayalakshmi Sethuraman4Department of Physics, Presidency College (Autonomous), University of MadrasSaama AI Research LabSaama AI Research LabPost-Graduate and Research Department of Physics, Centre for Nonlinear Science (CeNSc), Government College for Women (Autonomous)Department of Physics, Presidency College (Autonomous), University of MadrasAbstract Machine Learning methods are emerging as faster and efficient alternatives to numerical simulation techniques. The field of Scientific Computing has started adopting these data-driven approaches to faithfully model physical phenomena using scattered, noisy observations from coarse-grained grid-based simulations. In this paper, we investigate data-driven modelling of Bose-Einstein Condensates (BECs). In particular, we use Gaussian Processes (GPs) to model the ground state wave function of BECs as a function of scattering parameters from the dimensionless Gross Pitaveskii Equation (GPE). Experimental results illustrate the ability of GPs to accurately reproduce ground state wave functions using a limited number of data points from simulations. Consistent performance across different configurations of BECs, namely Scalar and Vectorial BECs generated under different potentials, including harmonic, double well and optical lattice potentials pronounces the versatility of our method. Comparison with existing data-driven models indicates that our model achieves similar accuracy with only a small fraction ( $$\frac{1}{50}$$ 1 50 th) of data points used by existing methods, in addition to modelling uncertainty from data. When used as a simulator post-training, our model generates ground state wave functions $$36 \times $$ 36 × faster than Trotter Suzuki, a numerical approximation technique that uses Imaginary time evolution. Our method is quite general; with minor changes it can be applied to similar quantum many-body problems.https://doi.org/10.1038/s41598-021-84336-0 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Tamil Arasan Bakthavatchalam Suriyadeepan Ramamoorthy Malaikannan Sankarasubbu Radha Ramaswamy Vijayalakshmi Sethuraman |
spellingShingle |
Tamil Arasan Bakthavatchalam Suriyadeepan Ramamoorthy Malaikannan Sankarasubbu Radha Ramaswamy Vijayalakshmi Sethuraman Bayesian Optimization of Bose-Einstein Condensates Scientific Reports |
author_facet |
Tamil Arasan Bakthavatchalam Suriyadeepan Ramamoorthy Malaikannan Sankarasubbu Radha Ramaswamy Vijayalakshmi Sethuraman |
author_sort |
Tamil Arasan Bakthavatchalam |
title |
Bayesian Optimization of Bose-Einstein Condensates |
title_short |
Bayesian Optimization of Bose-Einstein Condensates |
title_full |
Bayesian Optimization of Bose-Einstein Condensates |
title_fullStr |
Bayesian Optimization of Bose-Einstein Condensates |
title_full_unstemmed |
Bayesian Optimization of Bose-Einstein Condensates |
title_sort |
bayesian optimization of bose-einstein condensates |
publisher |
Nature Publishing Group |
series |
Scientific Reports |
issn |
2045-2322 |
publishDate |
2021-03-01 |
description |
Abstract Machine Learning methods are emerging as faster and efficient alternatives to numerical simulation techniques. The field of Scientific Computing has started adopting these data-driven approaches to faithfully model physical phenomena using scattered, noisy observations from coarse-grained grid-based simulations. In this paper, we investigate data-driven modelling of Bose-Einstein Condensates (BECs). In particular, we use Gaussian Processes (GPs) to model the ground state wave function of BECs as a function of scattering parameters from the dimensionless Gross Pitaveskii Equation (GPE). Experimental results illustrate the ability of GPs to accurately reproduce ground state wave functions using a limited number of data points from simulations. Consistent performance across different configurations of BECs, namely Scalar and Vectorial BECs generated under different potentials, including harmonic, double well and optical lattice potentials pronounces the versatility of our method. Comparison with existing data-driven models indicates that our model achieves similar accuracy with only a small fraction ( $$\frac{1}{50}$$ 1 50 th) of data points used by existing methods, in addition to modelling uncertainty from data. When used as a simulator post-training, our model generates ground state wave functions $$36 \times $$ 36 × faster than Trotter Suzuki, a numerical approximation technique that uses Imaginary time evolution. Our method is quite general; with minor changes it can be applied to similar quantum many-body problems. |
url |
https://doi.org/10.1038/s41598-021-84336-0 |
work_keys_str_mv |
AT tamilarasanbakthavatchalam bayesianoptimizationofboseeinsteincondensates AT suriyadeepanramamoorthy bayesianoptimizationofboseeinsteincondensates AT malaikannansankarasubbu bayesianoptimizationofboseeinsteincondensates AT radharamaswamy bayesianoptimizationofboseeinsteincondensates AT vijayalakshmisethuraman bayesianoptimizationofboseeinsteincondensates |
_version_ |
1724224645317525504 |