No Cover Image

Journal article 628 views 261 downloads

Extending machine learning classification capabilities with histogram reweighting

Dimitrios Bachtis, Gert Aarts Orcid Logo, Biagio Lucini Orcid Logo

Physical Review E, Volume: 102, Issue: 3

Swansea University Authors: Dimitrios Bachtis, Gert Aarts Orcid Logo, Biagio Lucini Orcid Logo

Check full text

DOI (Published version): 10.1103/physreve.102.033303

Abstract

We propose the use of Monte Carlo histogram reweighting to extrapolate predictions of machine learning methods. In our approach, we treat the output from a convolutional neural network as an observable in a statistical system, enabling its extrapolation over continuous ranges in parameter space. We...

Full description

Published in: Physical Review E
ISSN: 2470-0045 2470-0053
Published: American Physical Society (APS) 2020
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa55146
Tags: Add Tag
No Tags, Be the first to tag this record!
first_indexed 2020-09-09T16:28:07Z
last_indexed 2020-10-31T04:07:59Z
id cronfa55146
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2020-10-30T15:31:35.7009748</datestamp><bib-version>v2</bib-version><id>55146</id><entry>2020-09-09</entry><title>Extending machine learning classification capabilities with histogram reweighting</title><swanseaauthors><author><sid>91a311a58d3f8badc779f0ffa6d0ca3d</sid><firstname>Dimitrios</firstname><surname>Bachtis</surname><name>Dimitrios Bachtis</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>1ba0dad382dfe18348ec32fc65f3f3de</sid><ORCID>0000-0002-6038-3782</ORCID><firstname>Gert</firstname><surname>Aarts</surname><name>Gert Aarts</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>7e6fcfe060e07a351090e2a8aba363cf</sid><ORCID>0000-0001-8974-8266</ORCID><firstname>Biagio</firstname><surname>Lucini</surname><name>Biagio Lucini</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2020-09-09</date><deptcode>SPH</deptcode><abstract>We propose the use of Monte Carlo histogram reweighting to extrapolate predictions of machine learning methods. In our approach, we treat the output from a convolutional neural network as an observable in a statistical system, enabling its extrapolation over continuous ranges in parameter space. We demonstrate our proposal using the phase transition in the two-dimensional Ising model. By interpreting the output of the neural network as an order parameter, we explore connections with known observables in the system and investigate its scaling behavior. A finite-size scaling analysis is conducted based on quantities derived from the neural network that yields accurate estimates for the critical exponents and the critical temperature. The method improves the prospects of acquiring precision measurements from machine learning in physical systems without an order parameter and those where direct sampling in regions of parameter space might not be possible.</abstract><type>Journal Article</type><journal>Physical Review E</journal><volume>102</volume><journalNumber>3</journalNumber><paginationStart/><paginationEnd/><publisher>American Physical Society (APS)</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint>2470-0045</issnPrint><issnElectronic>2470-0053</issnElectronic><keywords/><publishedDay>9</publishedDay><publishedMonth>9</publishedMonth><publishedYear>2020</publishedYear><publishedDate>2020-09-09</publishedDate><doi>10.1103/physreve.102.033303</doi><url/><notes/><college>COLLEGE NANME</college><department>Physics</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>SPH</DepartmentCode><institution>Swansea University</institution><apcterm/><lastEdited>2020-10-30T15:31:35.7009748</lastEdited><Created>2020-09-09T17:26:57.8583742</Created><path><level id="1">Professional Services</level><level id="2">ISS - Uncategorised</level></path><authors><author><firstname>Dimitrios</firstname><surname>Bachtis</surname><order>1</order></author><author><firstname>Gert</firstname><surname>Aarts</surname><orcid>0000-0002-6038-3782</orcid><order>2</order></author><author><firstname>Biagio</firstname><surname>Lucini</surname><orcid>0000-0001-8974-8266</orcid><order>3</order></author></authors><documents><document><filename>55146__18133__0971da316d804d35b2a594f6835ef70b.pdf</filename><originalFilename>paper-histogram-final-version.pdf</originalFilename><uploaded>2020-09-09T18:14:33.6601047</uploaded><type>Output</type><contentLength>2240575</contentLength><contentType>application/pdf</contentType><version>Accepted Manuscript</version><cronfaStatus>true</cronfaStatus><copyrightCorrect>false</copyrightCorrect><language>eng</language></document></documents><OutputDurs/></rfc1807>
spelling 2020-10-30T15:31:35.7009748 v2 55146 2020-09-09 Extending machine learning classification capabilities with histogram reweighting 91a311a58d3f8badc779f0ffa6d0ca3d Dimitrios Bachtis Dimitrios Bachtis true false 1ba0dad382dfe18348ec32fc65f3f3de 0000-0002-6038-3782 Gert Aarts Gert Aarts true false 7e6fcfe060e07a351090e2a8aba363cf 0000-0001-8974-8266 Biagio Lucini Biagio Lucini true false 2020-09-09 SPH We propose the use of Monte Carlo histogram reweighting to extrapolate predictions of machine learning methods. In our approach, we treat the output from a convolutional neural network as an observable in a statistical system, enabling its extrapolation over continuous ranges in parameter space. We demonstrate our proposal using the phase transition in the two-dimensional Ising model. By interpreting the output of the neural network as an order parameter, we explore connections with known observables in the system and investigate its scaling behavior. A finite-size scaling analysis is conducted based on quantities derived from the neural network that yields accurate estimates for the critical exponents and the critical temperature. The method improves the prospects of acquiring precision measurements from machine learning in physical systems without an order parameter and those where direct sampling in regions of parameter space might not be possible. Journal Article Physical Review E 102 3 American Physical Society (APS) 2470-0045 2470-0053 9 9 2020 2020-09-09 10.1103/physreve.102.033303 COLLEGE NANME Physics COLLEGE CODE SPH Swansea University 2020-10-30T15:31:35.7009748 2020-09-09T17:26:57.8583742 Professional Services ISS - Uncategorised Dimitrios Bachtis 1 Gert Aarts 0000-0002-6038-3782 2 Biagio Lucini 0000-0001-8974-8266 3 55146__18133__0971da316d804d35b2a594f6835ef70b.pdf paper-histogram-final-version.pdf 2020-09-09T18:14:33.6601047 Output 2240575 application/pdf Accepted Manuscript true false eng
title Extending machine learning classification capabilities with histogram reweighting
spellingShingle Extending machine learning classification capabilities with histogram reweighting
Dimitrios Bachtis
Gert Aarts
Biagio Lucini
title_short Extending machine learning classification capabilities with histogram reweighting
title_full Extending machine learning classification capabilities with histogram reweighting
title_fullStr Extending machine learning classification capabilities with histogram reweighting
title_full_unstemmed Extending machine learning classification capabilities with histogram reweighting
title_sort Extending machine learning classification capabilities with histogram reweighting
author_id_str_mv 91a311a58d3f8badc779f0ffa6d0ca3d
1ba0dad382dfe18348ec32fc65f3f3de
7e6fcfe060e07a351090e2a8aba363cf
author_id_fullname_str_mv 91a311a58d3f8badc779f0ffa6d0ca3d_***_Dimitrios Bachtis
1ba0dad382dfe18348ec32fc65f3f3de_***_Gert Aarts
7e6fcfe060e07a351090e2a8aba363cf_***_Biagio Lucini
author Dimitrios Bachtis
Gert Aarts
Biagio Lucini
author2 Dimitrios Bachtis
Gert Aarts
Biagio Lucini
format Journal article
container_title Physical Review E
container_volume 102
container_issue 3
publishDate 2020
institution Swansea University
issn 2470-0045
2470-0053
doi_str_mv 10.1103/physreve.102.033303
publisher American Physical Society (APS)
college_str Professional Services
hierarchytype
hierarchy_top_id professionalservices
hierarchy_top_title Professional Services
hierarchy_parent_id professionalservices
hierarchy_parent_title Professional Services
department_str ISS - Uncategorised{{{_:::_}}}Professional Services{{{_:::_}}}ISS - Uncategorised
document_store_str 1
active_str 0
description We propose the use of Monte Carlo histogram reweighting to extrapolate predictions of machine learning methods. In our approach, we treat the output from a convolutional neural network as an observable in a statistical system, enabling its extrapolation over continuous ranges in parameter space. We demonstrate our proposal using the phase transition in the two-dimensional Ising model. By interpreting the output of the neural network as an order parameter, we explore connections with known observables in the system and investigate its scaling behavior. A finite-size scaling analysis is conducted based on quantities derived from the neural network that yields accurate estimates for the critical exponents and the critical temperature. The method improves the prospects of acquiring precision measurements from machine learning in physical systems without an order parameter and those where direct sampling in regions of parameter space might not be possible.
published_date 2020-09-09T04:09:09Z
_version_ 1763753643078582272
score 11.037581

Similar Items