Learning a reduced basis of dynamical systems using an autoencoder

Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to discover approximate finite-dimensional manifolds of two canonical partial differential equations. We test this method on the Kuramoto-Sivashinsky...

Full description

Saved in:
Bibliographic Details
Published in:Physical review. E Vol. 104; no. 3-1; p. 034202
Main Authors: Sondak, David, Protopapas, Pavlos
Format: Journal Article
Language:English
Published: 01.09.2021
ISSN:2470-0053, 2470-0053
Online Access:Get more information
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to discover approximate finite-dimensional manifolds of two canonical partial differential equations. We test this method on the Kuramoto-Sivashinsky (K-S), Korteweg-de Vries (KdV), and damped KdV equations. We show that the resulting optimal latent space of the K-S equation is consistent with the dimension of the inertial manifold. We then uncover a nonlinear basis representing the manifold of the latent space for the K-S equation. The results for the KdV equation show that it is more difficult to recover a reduced latent space, which is consistent with the truly infinite-dimensional dynamics of the KdV equation. In the case of the damped KdV equation, we find that the number of active dimensions decreases with increasing damping coefficient.Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to discover approximate finite-dimensional manifolds of two canonical partial differential equations. We test this method on the Kuramoto-Sivashinsky (K-S), Korteweg-de Vries (KdV), and damped KdV equations. We show that the resulting optimal latent space of the K-S equation is consistent with the dimension of the inertial manifold. We then uncover a nonlinear basis representing the manifold of the latent space for the K-S equation. The results for the KdV equation show that it is more difficult to recover a reduced latent space, which is consistent with the truly infinite-dimensional dynamics of the KdV equation. In the case of the damped KdV equation, we find that the number of active dimensions decreases with increasing damping coefficient.
AbstractList Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to discover approximate finite-dimensional manifolds of two canonical partial differential equations. We test this method on the Kuramoto-Sivashinsky (K-S), Korteweg-de Vries (KdV), and damped KdV equations. We show that the resulting optimal latent space of the K-S equation is consistent with the dimension of the inertial manifold. We then uncover a nonlinear basis representing the manifold of the latent space for the K-S equation. The results for the KdV equation show that it is more difficult to recover a reduced latent space, which is consistent with the truly infinite-dimensional dynamics of the KdV equation. In the case of the damped KdV equation, we find that the number of active dimensions decreases with increasing damping coefficient.Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to discover approximate finite-dimensional manifolds of two canonical partial differential equations. We test this method on the Kuramoto-Sivashinsky (K-S), Korteweg-de Vries (KdV), and damped KdV equations. We show that the resulting optimal latent space of the K-S equation is consistent with the dimension of the inertial manifold. We then uncover a nonlinear basis representing the manifold of the latent space for the K-S equation. The results for the KdV equation show that it is more difficult to recover a reduced latent space, which is consistent with the truly infinite-dimensional dynamics of the KdV equation. In the case of the damped KdV equation, we find that the number of active dimensions decreases with increasing damping coefficient.
Author Sondak, David
Protopapas, Pavlos
Author_xml – sequence: 1
  givenname: David
  surname: Sondak
  fullname: Sondak, David
– sequence: 2
  givenname: Pavlos
  surname: Protopapas
  fullname: Protopapas, Pavlos
BookMark eNpNjkFLAzEQRoNUsNb-Ai85etk6maZN9qilVmFBET2XdDLRlW1Wd3aF_ntFPXj63uHx-E7VKLeZlTo3MDMG5pcPrwd55M_1zICdwdwi4JEao3VQACzmo398oqYibwBgllA6g2N1XXHocp1fdNAdx4E46l2QWnSbdDzksK8pNFoO0vNe9CA_atZh6FvO1EbuztRxCo3w9G8n6vlm_bS6Lar7zd3qqioIPfaF3SXr6fsFuF0JQCUQJ_RLjIkZiSL6YG2iwJioJBdcLF3ygYzl6KjEibr47b537cfA0m_3tRA3TcjcDrLFhUdvrAeLX_xcU2M
CitedBy_id crossref_primary_10_3390_en18174747
crossref_primary_10_3847_1538_4357_ad1c55
crossref_primary_10_1016_j_jocs_2024_102471
crossref_primary_10_1017_jfm_2025_21
ContentType Journal Article
DBID 7X8
DOI 10.1103/PhysRevE.104.034202
DatabaseName MEDLINE - Academic
DatabaseTitle MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic
Database_xml – sequence: 1
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Physics
EISSN 2470-0053
GroupedDBID 3MX
53G
5VS
7X8
ABSSX
AEQTI
AFGMR
AGDNE
ALMA_UNASSIGNED_HOLDINGS
AUAIK
EBS
EJD
NPBMV
ROL
S7W
ID FETCH-LOGICAL-c282t-4bf48c00507b900c90cef2862dfee2ccd28a44fcae2fc9c7a7d97f8ac14ed7c92
IEDL.DBID 7X8
ISICitedReferencesCount 7
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000693323900005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2470-0053
IngestDate Thu Oct 02 10:26:21 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 3-1
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c282t-4bf48c00507b900c90cef2862dfee2ccd28a44fcae2fc9c7a7d97f8ac14ed7c92
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
PQID 2582814804
PQPubID 23479
ParticipantIDs proquest_miscellaneous_2582814804
PublicationCentury 2000
PublicationDate 2021-09-01
PublicationDateYYYYMMDD 2021-09-01
PublicationDate_xml – month: 09
  year: 2021
  text: 2021-09-01
  day: 01
PublicationDecade 2020
PublicationTitle Physical review. E
PublicationYear 2021
SSID ssj0001609712
Score 2.223786
Snippet Machine learning models have emerged as powerful tools in physics and engineering. In this work, we use an autoencoder with latent space penalization to...
SourceID proquest
SourceType Aggregation Database
StartPage 034202
Title Learning a reduced basis of dynamical systems using an autoencoder
URI https://www.proquest.com/docview/2582814804
Volume 104
WOSCitedRecordID wos000693323900005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1LSwMxEA5qFbz4Ft9E8BqbZuNuchKVFk9FRKG3kp1MpJdd2237-83sbineBM9JICSTyTw-vo-xO4tBa2cehPU-FzokWlhjU5FaSL1ODLqkEZvIhkMzGtm3tuBWtbDKlU-sHbUvgWrkXUX9nRi7S_34PRWkGkXd1VZCY5N1khjK0MPMRmZdY0mJIYkaCUqTvko0uJZ4qCeTLkEs33HZpzbnPTHh1aWV3w65_mUG-__d3wHba-NL_tQYxCHbwOKI7dQ4T6iO2XPLp_rFHZ8Rayt6Hj-yScXLwH2jTh-XN_zOFSdUfJxacLeYl0R56XF2wj4H_Y-XV9HKKAiI-dRc6DxoA0T0kuVWSrASMKiYyfiAqAC8Mk7rAA5VAAuZy7zNgnHQ0-gzsOqUbRVlgWeMx_zCq1TlqQMVB70DCfE-jTSonHTqnN2uDmcczZR6D67AclGN18dz8Yc5l2xXEXakxnJdsU6ITxGv2TYs55NqdlPf8g__r7JW
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Learning+a+reduced+basis+of+dynamical+systems+using+an+autoencoder&rft.jtitle=Physical+review.+E&rft.au=Sondak%2C+David&rft.au=Protopapas%2C+Pavlos&rft.date=2021-09-01&rft.issn=2470-0053&rft.eissn=2470-0053&rft.volume=104&rft.issue=3-1&rft.spage=034202&rft_id=info:doi/10.1103%2FPhysRevE.104.034202&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2470-0053&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2470-0053&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2470-0053&client=summon