Hierarchical optimization for neutron scattering problems

Feng Bao; Rick Archibald; Dipanshu Bansal; Olivier Delaire

doi:10.1016/j.jcp.2016.03.017

Hierarchical optimization for neutron scattering problems

Feng Bao, Rick Archibald, Dipanshu Bansal, Olivier Delaire

Data Analysis and Machine Learning

Research output: Contribution to journal › Article › peer-review

3 Scopus citations

Abstract

We present a scalable optimization method for neutron scattering problems that determines confidence regions of simulation parameters in lattice dynamics models used to fit neutron scattering data for crystalline solids. The method uses physics-based hierarchical dimension reduction in both the computational simulation domain and the parameter space. We demonstrate for silicon that after a few iterations the method converges to parameters values (interatomic force-constants) computed with density functional theory simulations.

Original language	English
Pages (from-to)	39-51
Number of pages	13
Journal	Journal of Computational Physics
Volume	315
DOIs	https://doi.org/10.1016/j.jcp.2016.03.017
State	Published - Jun 15 2016

Funding

Acknowledges support by the U.S. Department of Energy , Office of Science , Basic Energy Sciences, Materials Sciences and Engineering Division, through the Office of Science Early Career Research Program and Advanced Scientific Computing Research , through the ACUMEN project.

Funders	Funder number
U.S. Department of Energy
Office of Science
Basic Energy Sciences
Division of Materials Sciences and Engineering

Keywords

Confidence distribution
Global optimization
Model reduction
Neutron scattering
Stochastic

Access to Document

10.1016/j.jcp.2016.03.017

Cite this

@article{dc2c22a39dda4a1e80fa2ea4893a87a7,

title = "Hierarchical optimization for neutron scattering problems",

abstract = "We present a scalable optimization method for neutron scattering problems that determines confidence regions of simulation parameters in lattice dynamics models used to fit neutron scattering data for crystalline solids. The method uses physics-based hierarchical dimension reduction in both the computational simulation domain and the parameter space. We demonstrate for silicon that after a few iterations the method converges to parameters values (interatomic force-constants) computed with density functional theory simulations.",

keywords = "Confidence distribution, Global optimization, Model reduction, Neutron scattering, Stochastic",

author = "Feng Bao and Rick Archibald and Dipanshu Bansal and Olivier Delaire",

note = "Publisher Copyright: {\textcopyright} 2016.",

year = "2016",

month = jun,

day = "15",

doi = "10.1016/j.jcp.2016.03.017",

language = "English",

volume = "315",

pages = "39--51",

journal = "Journal of Computational Physics",

issn = "0021-9991",

publisher = "Elsevier",

}

TY - JOUR

T1 - Hierarchical optimization for neutron scattering problems

AU - Bao, Feng

AU - Archibald, Rick

AU - Bansal, Dipanshu

AU - Delaire, Olivier

PY - 2016/6/15

Y1 - 2016/6/15

N2 - We present a scalable optimization method for neutron scattering problems that determines confidence regions of simulation parameters in lattice dynamics models used to fit neutron scattering data for crystalline solids. The method uses physics-based hierarchical dimension reduction in both the computational simulation domain and the parameter space. We demonstrate for silicon that after a few iterations the method converges to parameters values (interatomic force-constants) computed with density functional theory simulations.

AB - We present a scalable optimization method for neutron scattering problems that determines confidence regions of simulation parameters in lattice dynamics models used to fit neutron scattering data for crystalline solids. The method uses physics-based hierarchical dimension reduction in both the computational simulation domain and the parameter space. We demonstrate for silicon that after a few iterations the method converges to parameters values (interatomic force-constants) computed with density functional theory simulations.

KW - Confidence distribution

KW - Global optimization

KW - Model reduction

KW - Neutron scattering

KW - Stochastic

UR - http://www.scopus.com/inward/record.url?scp=84962286675&partnerID=8YFLogxK

U2 - 10.1016/j.jcp.2016.03.017

DO - 10.1016/j.jcp.2016.03.017

M3 - Article

AN - SCOPUS:84962286675

SN - 0021-9991

VL - 315

SP - 39

EP - 51

JO - Journal of Computational Physics

JF - Journal of Computational Physics

ER -

Hierarchical optimization for neutron scattering problems

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this