A machine learning inversion scheme for determining interaction from scattering

Ming Ching Chang; Chi Huan Tung; Shou Yi Chang; Jan Michael Carrillo; Yangyang Wang; Bobby G. Sumpter; Guan Rong Huang; Changwoo Do; Wei Ren Chen

doi:10.1038/s42005-021-00778-y

A machine learning inversion scheme for determining interaction from scattering

Ming Ching Chang, Chi Huan Tung, Shou Yi Chang, Jan Michael Carrillo, Yangyang Wang, Bobby G. Sumpter, Guan Rong Huang, Changwoo Do, Wei Ren Chen

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

15 Scopus citations

Abstract

Small angle scattering techniques have now been routinely used to quantitatively determine the potential of mean force in colloidal suspensions. However the numerical accuracy of data interpretation is often compounded by the approximations adopted by liquid state analytical theories. To circumvent this long standing issue, here we outline a machine learning strategy for determining the effective interaction in the condensed phases of matter using scattering. Via a case study of colloidal suspensions, we show that the effective potential can be probabilistically inferred from the scattering spectra without any restriction imposed by model assumptions. Comparisons to existing parametric approaches demonstrate the superior performance of this method in accuracy, efficiency, and applicability. This method can effectively enable quantification of interaction in highly correlated systems using scattering and diffraction experiments.

Original language	English
Article number	46
Journal	Communications Physics
Volume	5
Issue number	1
DOIs	https://doi.org/10.1038/s42005-021-00778-y
State	Published - Dec 2022

Funding

We thank Y. Shinohara, T. Egami, P. Falus, L. Porcar, and Y. Liu for helpful discussions. This research was performed at the Spallation Neutron Source and the Center for Nanophase Materials Sciences, which are US Department of Energy (DOE) Office of Science User Facilities operated by ORNL. MD simulations used resources of the Oak Ridge Leadership Computing Facility, which is supported by DOE Office of Science under Contract DE-AC05-00OR22725. C.-H.T. thanks the financial support from the Shull Wollan Center during his stay at ORNL. Y.W. is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Early Career Research Program Award KC0402010, under Contract DE-AC05-00OR22725. B.G.S. acknowledges support by by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences Data, Artificial Intelligence and Machine Learning at DOE Scientific User Facilities Program under Award Number 34532. M.-C.C. thanks the support provided by the University at Albany—SUNY.

Funders	Funder number
Center for Nanophase Materials Sciences
Office of Basic Energy Sciences Data, Artificial Intelligence and Machine Learning
Shull Wollan Center
U.S. Department of Energy	34532
Office of Science	DE-AC05-00OR22725
Basic Energy Sciences	KC0402010
Oak Ridge National Laboratory
State University of New York
University at Albany

Access to Document

10.1038/s42005-021-00778-y

Cite this

@article{f9a2f66748f14e8185f5d5babf133b75,

title = "A machine learning inversion scheme for determining interaction from scattering",

abstract = "Small angle scattering techniques have now been routinely used to quantitatively determine the potential of mean force in colloidal suspensions. However the numerical accuracy of data interpretation is often compounded by the approximations adopted by liquid state analytical theories. To circumvent this long standing issue, here we outline a machine learning strategy for determining the effective interaction in the condensed phases of matter using scattering. Via a case study of colloidal suspensions, we show that the effective potential can be probabilistically inferred from the scattering spectra without any restriction imposed by model assumptions. Comparisons to existing parametric approaches demonstrate the superior performance of this method in accuracy, efficiency, and applicability. This method can effectively enable quantification of interaction in highly correlated systems using scattering and diffraction experiments.",

author = "Chang, {Ming Ching} and Tung, {Chi Huan} and Chang, {Shou Yi} and Carrillo, {Jan Michael} and Yangyang Wang and Sumpter, {Bobby G.} and Huang, {Guan Rong} and Changwoo Do and Chen, {Wei Ren}",

note = "Publisher Copyright: {\textcopyright} 2022, UT-Battelle, LLC.",

year = "2022",

month = dec,

doi = "10.1038/s42005-021-00778-y",

language = "English",

volume = "5",

journal = "Communications Physics",

issn = "2399-3650",

publisher = "Nature Research",

number = "1",

}

TY - JOUR

T1 - A machine learning inversion scheme for determining interaction from scattering

AU - Chang, Ming Ching

AU - Tung, Chi Huan

AU - Chang, Shou Yi

AU - Carrillo, Jan Michael

AU - Wang, Yangyang

AU - Sumpter, Bobby G.

AU - Huang, Guan Rong

AU - Do, Changwoo

AU - Chen, Wei Ren

PY - 2022/12

Y1 - 2022/12

N2 - Small angle scattering techniques have now been routinely used to quantitatively determine the potential of mean force in colloidal suspensions. However the numerical accuracy of data interpretation is often compounded by the approximations adopted by liquid state analytical theories. To circumvent this long standing issue, here we outline a machine learning strategy for determining the effective interaction in the condensed phases of matter using scattering. Via a case study of colloidal suspensions, we show that the effective potential can be probabilistically inferred from the scattering spectra without any restriction imposed by model assumptions. Comparisons to existing parametric approaches demonstrate the superior performance of this method in accuracy, efficiency, and applicability. This method can effectively enable quantification of interaction in highly correlated systems using scattering and diffraction experiments.

AB - Small angle scattering techniques have now been routinely used to quantitatively determine the potential of mean force in colloidal suspensions. However the numerical accuracy of data interpretation is often compounded by the approximations adopted by liquid state analytical theories. To circumvent this long standing issue, here we outline a machine learning strategy for determining the effective interaction in the condensed phases of matter using scattering. Via a case study of colloidal suspensions, we show that the effective potential can be probabilistically inferred from the scattering spectra without any restriction imposed by model assumptions. Comparisons to existing parametric approaches demonstrate the superior performance of this method in accuracy, efficiency, and applicability. This method can effectively enable quantification of interaction in highly correlated systems using scattering and diffraction experiments.

UR - http://www.scopus.com/inward/record.url?scp=85125634578&partnerID=8YFLogxK

U2 - 10.1038/s42005-021-00778-y

DO - 10.1038/s42005-021-00778-y

M3 - Article

AN - SCOPUS:85125634578

SN - 2399-3650

VL - 5

JO - Communications Physics

JF - Communications Physics

IS - 1

M1 - 46

ER -

A machine learning inversion scheme for determining interaction from scattering

Abstract

Funding

Access to Document

Other files and links

Fingerprint

Cite this