Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making

This paper advocates integrating two decades-old tools, namely, input-space bisection and backpropagation, in support of design decision making. Either technique independently could offer some improvements in efficiency and reduced computational costs; however, taken in concert their effects could b...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Cowan, M.A., Butler, C.D.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 611
container_issue
container_start_page 604
container_title
container_volume
creator Cowan, M.A.
Butler, C.D.
description This paper advocates integrating two decades-old tools, namely, input-space bisection and backpropagation, in support of design decision making. Either technique independently could offer some improvements in efficiency and reduced computational costs; however, taken in concert their effects could be greatly amplified. Traditional design decisions have often relied too heavily upon expensive, time-consuming HPC runs. This paper demonstrates from three disparate cases that a carefully chosen input space serving as training instances for a backpropagation algorithm can often perform within acceptable error bounds for the test cases, while providing a much quicker response to decision makers. First, we introduce the problem domain, stressing the inadequacies and the inefficiencies of using too highly resolved parameter variations in batch HPC runs. We then provide a step-by-step overview of our surrogate model approach. We describe the similarities of the requirements for running both numerical solvers and artificial neural networks. Our method entails using an input-space bisection to reduce (in both the spatial and temporal domains) the output of the numerical solver, the results of which will serve as training cases for the backpropagation algorithm. The technique is demonstrated in three test cases, two of which are built to predict behavior at a fixed point in time, while the other, more ambitiously, tackles contaminant trace in hydraulic flow over multiple time-steps. Finally, we summarize some of the benefits this approach offers and outline directions for future research.
doi_str_mv 10.1109/IJCNN.2009.5178629
format Conference Proceeding
fullrecord <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_5178629</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>5178629</ieee_id><sourcerecordid>5178629</sourcerecordid><originalsourceid>FETCH-ieee_primary_51786293</originalsourceid><addsrcrecordid>eNp9z71Ow0AMB_DjS6JAXwAWv0DCXXJJm7kCFYZOILFV7sUNR3MfihMJJl6dUAojXv6Sf7ItC3GtZKqUrG4fHherVZpJWaWFms3LrDoSF0pnWudFkZfHYpKpUiVay9nJH-j5y-kv5FV-LqbMb3KsPcqJ-Hxm6xuwPg59whENwcYymd4GD-hr2KDZxS5EbPCnx4DgqH8NNfQByA0t9gQmuNjSO_jBUWcNtsB2T-MMwzZ0UBPbxo9hLH8vcrgbL1-Jsy22TNNDXoqb-7unxTKxRLSOnXXYfawPD-f_6xedCldQ</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Cowan, M.A. ; Butler, C.D.</creator><creatorcontrib>Cowan, M.A. ; Butler, C.D.</creatorcontrib><description>This paper advocates integrating two decades-old tools, namely, input-space bisection and backpropagation, in support of design decision making. Either technique independently could offer some improvements in efficiency and reduced computational costs; however, taken in concert their effects could be greatly amplified. Traditional design decisions have often relied too heavily upon expensive, time-consuming HPC runs. This paper demonstrates from three disparate cases that a carefully chosen input space serving as training instances for a backpropagation algorithm can often perform within acceptable error bounds for the test cases, while providing a much quicker response to decision makers. First, we introduce the problem domain, stressing the inadequacies and the inefficiencies of using too highly resolved parameter variations in batch HPC runs. We then provide a step-by-step overview of our surrogate model approach. We describe the similarities of the requirements for running both numerical solvers and artificial neural networks. Our method entails using an input-space bisection to reduce (in both the spatial and temporal domains) the output of the numerical solver, the results of which will serve as training cases for the backpropagation algorithm. The technique is demonstrated in three test cases, two of which are built to predict behavior at a fixed point in time, while the other, more ambitiously, tackles contaminant trace in hydraulic flow over multiple time-steps. Finally, we summarize some of the benefits this approach offers and outline directions for future research.</description><identifier>ISSN: 2161-4393</identifier><identifier>ISBN: 142443548X</identifier><identifier>ISBN: 9781424435487</identifier><identifier>EISSN: 2161-4407</identifier><identifier>EISBN: 1424435536</identifier><identifier>EISBN: 9781424435531</identifier><identifier>DOI: 10.1109/IJCNN.2009.5178629</identifier><language>eng</language><publisher>IEEE</publisher><subject>Artificial neural networks ; Backpropagation algorithms ; Computational modeling ; Decision making ; Design engineering ; Information technology ; Laboratories ; Numerical simulation ; Research and development ; Testing</subject><ispartof>2009 International Joint Conference on Neural Networks, 2009, p.604-611</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/5178629$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2056,27924,54919</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/5178629$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Cowan, M.A.</creatorcontrib><creatorcontrib>Butler, C.D.</creatorcontrib><title>Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making</title><title>2009 International Joint Conference on Neural Networks</title><addtitle>IJCNN</addtitle><description>This paper advocates integrating two decades-old tools, namely, input-space bisection and backpropagation, in support of design decision making. Either technique independently could offer some improvements in efficiency and reduced computational costs; however, taken in concert their effects could be greatly amplified. Traditional design decisions have often relied too heavily upon expensive, time-consuming HPC runs. This paper demonstrates from three disparate cases that a carefully chosen input space serving as training instances for a backpropagation algorithm can often perform within acceptable error bounds for the test cases, while providing a much quicker response to decision makers. First, we introduce the problem domain, stressing the inadequacies and the inefficiencies of using too highly resolved parameter variations in batch HPC runs. We then provide a step-by-step overview of our surrogate model approach. We describe the similarities of the requirements for running both numerical solvers and artificial neural networks. Our method entails using an input-space bisection to reduce (in both the spatial and temporal domains) the output of the numerical solver, the results of which will serve as training cases for the backpropagation algorithm. The technique is demonstrated in three test cases, two of which are built to predict behavior at a fixed point in time, while the other, more ambitiously, tackles contaminant trace in hydraulic flow over multiple time-steps. Finally, we summarize some of the benefits this approach offers and outline directions for future research.</description><subject>Artificial neural networks</subject><subject>Backpropagation algorithms</subject><subject>Computational modeling</subject><subject>Decision making</subject><subject>Design engineering</subject><subject>Information technology</subject><subject>Laboratories</subject><subject>Numerical simulation</subject><subject>Research and development</subject><subject>Testing</subject><issn>2161-4393</issn><issn>2161-4407</issn><isbn>142443548X</isbn><isbn>9781424435487</isbn><isbn>1424435536</isbn><isbn>9781424435531</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2009</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNp9z71Ow0AMB_DjS6JAXwAWv0DCXXJJm7kCFYZOILFV7sUNR3MfihMJJl6dUAojXv6Sf7ItC3GtZKqUrG4fHherVZpJWaWFms3LrDoSF0pnWudFkZfHYpKpUiVay9nJH-j5y-kv5FV-LqbMb3KsPcqJ-Hxm6xuwPg59whENwcYymd4GD-hr2KDZxS5EbPCnx4DgqH8NNfQByA0t9gQmuNjSO_jBUWcNtsB2T-MMwzZ0UBPbxo9hLH8vcrgbL1-Jsy22TNNDXoqb-7unxTKxRLSOnXXYfawPD-f_6xedCldQ</recordid><startdate>200906</startdate><enddate>200906</enddate><creator>Cowan, M.A.</creator><creator>Butler, C.D.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>200906</creationdate><title>Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making</title><author>Cowan, M.A. ; Butler, C.D.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-ieee_primary_51786293</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2009</creationdate><topic>Artificial neural networks</topic><topic>Backpropagation algorithms</topic><topic>Computational modeling</topic><topic>Decision making</topic><topic>Design engineering</topic><topic>Information technology</topic><topic>Laboratories</topic><topic>Numerical simulation</topic><topic>Research and development</topic><topic>Testing</topic><toplevel>online_resources</toplevel><creatorcontrib>Cowan, M.A.</creatorcontrib><creatorcontrib>Butler, C.D.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Cowan, M.A.</au><au>Butler, C.D.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making</atitle><btitle>2009 International Joint Conference on Neural Networks</btitle><stitle>IJCNN</stitle><date>2009-06</date><risdate>2009</risdate><spage>604</spage><epage>611</epage><pages>604-611</pages><issn>2161-4393</issn><eissn>2161-4407</eissn><isbn>142443548X</isbn><isbn>9781424435487</isbn><eisbn>1424435536</eisbn><eisbn>9781424435531</eisbn><abstract>This paper advocates integrating two decades-old tools, namely, input-space bisection and backpropagation, in support of design decision making. Either technique independently could offer some improvements in efficiency and reduced computational costs; however, taken in concert their effects could be greatly amplified. Traditional design decisions have often relied too heavily upon expensive, time-consuming HPC runs. This paper demonstrates from three disparate cases that a carefully chosen input space serving as training instances for a backpropagation algorithm can often perform within acceptable error bounds for the test cases, while providing a much quicker response to decision makers. First, we introduce the problem domain, stressing the inadequacies and the inefficiencies of using too highly resolved parameter variations in batch HPC runs. We then provide a step-by-step overview of our surrogate model approach. We describe the similarities of the requirements for running both numerical solvers and artificial neural networks. Our method entails using an input-space bisection to reduce (in both the spatial and temporal domains) the output of the numerical solver, the results of which will serve as training cases for the backpropagation algorithm. The technique is demonstrated in three test cases, two of which are built to predict behavior at a fixed point in time, while the other, more ambitiously, tackles contaminant trace in hydraulic flow over multiple time-steps. Finally, we summarize some of the benefits this approach offers and outline directions for future research.</abstract><pub>IEEE</pub><doi>10.1109/IJCNN.2009.5178629</doi></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 2161-4393
ispartof 2009 International Joint Conference on Neural Networks, 2009, p.604-611
issn 2161-4393
2161-4407
language eng
recordid cdi_ieee_primary_5178629
source IEEE Electronic Library (IEL) Conference Proceedings
subjects Artificial neural networks
Backpropagation algorithms
Computational modeling
Decision making
Design engineering
Information technology
Laboratories
Numerical simulation
Research and development
Testing
title Using input-space bisection and backpropagation as a method to emulate complex numerical simulations for design decision making
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-12T10%3A29%3A11IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Using%20input-space%20bisection%20and%20backpropagation%20as%20a%20method%20to%20emulate%20complex%20numerical%20simulations%20for%20design%20decision%20making&rft.btitle=2009%20International%20Joint%20Conference%20on%20Neural%20Networks&rft.au=Cowan,%20M.A.&rft.date=2009-06&rft.spage=604&rft.epage=611&rft.pages=604-611&rft.issn=2161-4393&rft.eissn=2161-4407&rft.isbn=142443548X&rft.isbn_list=9781424435487&rft_id=info:doi/10.1109/IJCNN.2009.5178629&rft_dat=%3Cieee_6IE%3E5178629%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&rft.eisbn=1424435536&rft.eisbn_list=9781424435531&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=5178629&rfr_iscdi=true