An Inexact Sequential Quadratic Programming Method for Learning and Control of Recurrent Neural Networks

This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal) control stage. We present an inexact sequential quadratic programming framework for recurrent neural network learning (iSQPRL) for solving the i...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE transaction on neural networks and learning systems Ročník 36; číslo 2; s. 2762 - 2776
Hlavní autori: Adeoye, Adeyemi D., Bemporad, Alberto
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States IEEE 01.02.2025
Predmet:
ISSN:2162-237X, 2162-2388, 2162-2388
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal) control stage. We present an inexact sequential quadratic programming framework for recurrent neural network learning (iSQPRL) for solving the identification stage of the POMDP, in which the true system is approximated by a recurrent neural network (RNN) with dynamically consistent overshooting (DCRNN). We formulate the learning problem as a constrained optimization problem and study the quadratic programming (QP) subproblem with a convergence analysis under a restarted Krylov-subspace iterative scheme that implicitly exploits the structure of the associated Karush-Kuhn-Tucker (KKT) subsystem. In the control stage, where a feedforward neural network (FNN) controller is designed on top of the RNN model, we adapt a generalized Gauss-Newton (GGN) algorithm that exploits useful approximations to the curvature terms of the training data and selects its mini-batch step size using a known property of some regularization function. Simulation results are provided to demonstrate the effectiveness of our approach.
AbstractList This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal) control stage. We present an inexact sequential quadratic programming framework for recurrent neural network learning (iSQPRL) for solving the identification stage of the POMDP, in which the true system is approximated by a recurrent neural network (RNN) with dynamically consistent overshooting (DCRNN). We formulate the learning problem as a constrained optimization problem and study the quadratic programming (QP) subproblem with a convergence analysis under a restarted Krylov-subspace iterative scheme that implicitly exploits the structure of the associated Karush-Kuhn-Tucker (KKT) subsystem. In the control stage, where a feedforward neural network (FNN) controller is designed on top of the RNN model, we adapt a generalized Gauss-Newton (GGN) algorithm that exploits useful approximations to the curvature terms of the training data and selects its mini-batch step size using a known property of some regularization function. Simulation results are provided to demonstrate the effectiveness of our approach.This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal) control stage. We present an inexact sequential quadratic programming framework for recurrent neural network learning (iSQPRL) for solving the identification stage of the POMDP, in which the true system is approximated by a recurrent neural network (RNN) with dynamically consistent overshooting (DCRNN). We formulate the learning problem as a constrained optimization problem and study the quadratic programming (QP) subproblem with a convergence analysis under a restarted Krylov-subspace iterative scheme that implicitly exploits the structure of the associated Karush-Kuhn-Tucker (KKT) subsystem. In the control stage, where a feedforward neural network (FNN) controller is designed on top of the RNN model, we adapt a generalized Gauss-Newton (GGN) algorithm that exploits useful approximations to the curvature terms of the training data and selects its mini-batch step size using a known property of some regularization function. Simulation results are provided to demonstrate the effectiveness of our approach.
This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal) control stage. We present an inexact sequential quadratic programming framework for recurrent neural network learning (iSQPRL) for solving the identification stage of the POMDP, in which the true system is approximated by a recurrent neural network (RNN) with dynamically consistent overshooting (DCRNN). We formulate the learning problem as a constrained optimization problem and study the quadratic programming (QP) subproblem with a convergence analysis under a restarted Krylov-subspace iterative scheme that implicitly exploits the structure of the associated Karush-Kuhn-Tucker (KKT) subsystem. In the control stage, where a feedforward neural network (FNN) controller is designed on top of the RNN model, we adapt a generalized Gauss-Newton (GGN) algorithm that exploits useful approximations to the curvature terms of the training data and selects its mini-batch step size using a known property of some regularization function. Simulation results are provided to demonstrate the effectiveness of our approach.
Author Adeoye, Adeyemi D.
Bemporad, Alberto
Author_xml – sequence: 1
  givenname: Adeyemi D.
  orcidid: 0000-0001-7048-0984
  surname: Adeoye
  fullname: Adeoye, Adeyemi D.
  email: adeyemi.adeoye@imtlucca.it
  organization: IMT School for Advanced Studies Lucca, Lucca, Italy
– sequence: 2
  givenname: Alberto
  orcidid: 0000-0001-6761-0856
  surname: Bemporad
  fullname: Bemporad, Alberto
  organization: IMT School for Advanced Studies Lucca, Lucca, Italy
BackLink https://www.ncbi.nlm.nih.gov/pubmed/38294918$$D View this record in MEDLINE/PubMed
BookMark eNp9kUtLxDAUhYMovv-AiGTpZsa82kmWMviCcXyDu5KmN1ptE01S1H9vxxlFXJjNDZfvnJBzNtCy8w4Q2qFkSClRB7fT6eRmyAgTQ84zIbNsCa0zmrMB41Iu_9xH92toO8Yn0p-cZLlQq2iNS6aEonIdPR46fObgXZuEb-C1A5dq3eCrTldBp9rgy-Afgm7b2j3gc0iPvsLWBzwBHdxsp12Fx96l4BvsLb4G04XQu-ApdKF3mkJ68-E5bqEVq5sI24u5ie6Oj27Hp4PJxcnZ-HAyMDyXacBsaRixpDImp7IaqZwJkwO3nEhVGk6sElrxTFmZQWYrLpgsCSkZo5kZyYpvov2570vw_XdiKto6Gmga7cB3sWCKUUpHhIse3VugXdlCVbyEutXho_hOpwfYHDDBxxjA_iCUFLMWiq8WilkLxaKFXiT_iEyd-ihnGem6-V-6O5fWAPDrLUElEZx_AvBclS4
CODEN ITNNAL
CitedBy_id crossref_primary_10_3390_math13111766
crossref_primary_10_1109_TASE_2025_3576586
crossref_primary_10_1007_s00521_024_10743_9
Cites_doi 10.23919/ACC.1992.4792127
10.1109/tnnls.2021.3109565
10.2307/2004873
10.1109/TNNLS.2021.3105818
10.1109/ADPRL.2007.368182
10.1007/BF01588967
10.21105/joss.00598
10.1109/TNNLS.2014.2361267
10.1109/72.279191
10.1109/TNNLS.2019.2953622
10.1109/IJCNN.1990.137723
10.21236/ADA164453
10.1109/IJCNN.1999.832595
10.1109/72.846741
10.1016/S0893-6080(98)00116-6
10.1016/j.compchemeng.2016.04.026
10.1137/0728063
10.1109/IJCNN.1992.227335
10.1109/TNN.2011.2109737
10.1137/060674004
10.1137/0907058
10.1016/j.na.2005.02.015
10.1109/tac.2022.3222750
10.1142/S0218488598000094
10.1007/978-3-540-89722-4_11
10.1007/978-3-319-91578-4
10.2140/pjm.1966.16.1
10.1109/IJCNN.1999.832591
10.1137/1.9781611970791
10.1016/S0925-2312(98)00104-0
10.1109/tnnls.2021.3109953
10.1109/tnn.1998.712192
10.1007/b98874
10.1109/TNNLS.2017.2741598
10.1109/tnnls.2022.3151412
10.1109/72.279181
10.1007/s10107-008-0248-3
10.7551/mitpress/4977.003.0010
10.21105/joss.00602
10.1017/S0962492904000212
10.1016/j.automatica.2010.06.021
10.1109/IJCNN.2007.4370969
10.1090/S0025-5718-1970-0274029-X
10.1093/imamat/6.1.76
10.1093/comjnl/13.3.317
10.1137/0719025
10.1007/BFb0067703
10.1016/j.automatica.2023.111183
10.1007/s10589-023-00502-2
10.1038/nature14539
10.1109/ICCV.2015.123
ContentType Journal Article
DBID 97E
ESBDL
RIA
RIE
AAYXX
CITATION
NPM
7X8
DOI 10.1109/TNNLS.2024.3354855
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005-present
IEEE Open Access Journals
IEEE All-Society Periodicals Package (ASPP) 1998-Present
IEEE Electronic Library (IEL)
CrossRef
PubMed
MEDLINE - Academic
DatabaseTitle CrossRef
PubMed
MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic
PubMed

Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 3
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EISSN 2162-2388
EndPage 2776
ExternalDocumentID 38294918
10_1109_TNNLS_2024_3354855
10418043
Genre orig-research
Journal Article
GroupedDBID 0R~
4.4
5VS
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABQJQ
ABVLG
ACIWK
ACPRK
AENEX
AFRAH
AGQYO
AGSQL
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
EBS
EJD
ESBDL
IFIPE
IPLJI
JAVBF
M43
MS~
O9-
OCL
PQQKQ
RIA
RIE
RNS
AAYXX
CITATION
NPM
RIG
7X8
ID FETCH-LOGICAL-c368t-2fbc20f0dcc618d79624c6e3f3089bc30f94a9359f85e5fd3428b00b2215c78d3
IEDL.DBID RIE
ISICitedReferencesCount 6
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001157917100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2162-237X
2162-2388
IngestDate Wed Oct 01 13:17:36 EDT 2025
Mon Jul 21 06:03:48 EDT 2025
Sat Nov 29 01:40:30 EST 2025
Tue Nov 18 22:31:03 EST 2025
Wed Aug 27 01:52:59 EDT 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed false
IsScholarly true
Issue 2
Language English
License https://creativecommons.org/licenses/by/4.0/legalcode
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c368t-2fbc20f0dcc618d79624c6e3f3089bc30f94a9359f85e5fd3428b00b2215c78d3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0001-6761-0856
0000-0001-7048-0984
OpenAccessLink https://ieeexplore.ieee.org/document/10418043
PMID 38294918
PQID 2921117034
PQPubID 23479
PageCount 15
ParticipantIDs pubmed_primary_38294918
proquest_miscellaneous_2921117034
crossref_citationtrail_10_1109_TNNLS_2024_3354855
crossref_primary_10_1109_TNNLS_2024_3354855
ieee_primary_10418043
PublicationCentury 2000
PublicationDate 2025-02-01
PublicationDateYYYYMMDD 2025-02-01
PublicationDate_xml – month: 02
  year: 2025
  text: 2025-02-01
  day: 01
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle IEEE transaction on neural networks and learning systems
PublicationTitleAbbrev TNNLS
PublicationTitleAlternate IEEE Trans Neural Netw Learn Syst
PublicationYear 2025
Publisher IEEE
Publisher_xml – name: IEEE
References Schmidhuber (ref7); 3
ref13
ref57
ref12
ref56
ref15
Schäfer (ref59) 2008
ref53
ref52
Innes (ref65) 2018
ref10
ref17
ref16
Li (ref30) 2017
ref19
ref18
Park (ref54) 1975
Adeoye (ref68) 2023
ref51
ref50
ref46
ref45
ref42
ref44
Bonnans (ref48) 2006
ref43
Choquet (ref47) 1993
ref49
ref8
ref9
Ha (ref4); 31
ref3
ref6
ref40
Haykin (ref1) 2009
Hochreiter (ref11) 1998; 6
ref35
ref34
ref37
ref36
ref31
ref33
ref32
Wills (ref58) 2018
ref2
ref39
ref38
Kingma (ref55) 2014
Duan (ref61)
Schmidhuber (ref5) 1990
ref24
Powell (ref41) 1978
ref23
ref67
ref26
Martens (ref14)
ref25
Revels (ref64) 2016
ref20
ref63
ref22
ref66
ref21
ref28
ref27
ref29
ref60
ref62
References_xml – ident: ref20
  doi: 10.23919/ACC.1992.4792127
– ident: ref27
  doi: 10.1109/tnnls.2021.3109565
– ident: ref37
  doi: 10.2307/2004873
– year: 1993
  ident: ref47
  article-title: Some convergence results for the Newton-GMRES algorithm
– ident: ref63
  doi: 10.1109/TNNLS.2021.3105818
– ident: ref8
  doi: 10.1109/ADPRL.2007.368182
– year: 2018
  ident: ref58
  article-title: Stochastic quasi-Newton with adaptive step lengths for large-scale problems
  publication-title: arXiv:1802.04310
– ident: ref40
  doi: 10.1007/BF01588967
– volume-title: Numerical Optimization: Theoretical and Practical Aspects
  year: 2006
  ident: ref48
– year: 2017
  ident: ref30
  article-title: Maximum principle based algorithms for deep learning
  publication-title: arXiv:1710.09513
– ident: ref67
  doi: 10.21105/joss.00598
– ident: ref17
  doi: 10.1109/TNNLS.2014.2361267
– ident: ref22
  doi: 10.1109/72.279191
– ident: ref26
  doi: 10.1109/TNNLS.2019.2953622
– ident: ref6
  doi: 10.1109/IJCNN.1990.137723
– year: 1990
  ident: ref5
  article-title: Making the world differentiable: On using selfsupervised fully recurrent neural networks for dynamic reinforcement learning and planning in non-stationary environments
– ident: ref31
  doi: 10.21236/ADA164453
– ident: ref15
  doi: 10.1109/IJCNN.1999.832595
– year: 2018
  ident: ref65
  article-title: Fashionable modelling with flux
  publication-title: arXiv:1811.01457
– ident: ref32
  doi: 10.1109/72.846741
– ident: ref56
  doi: 10.1016/S0893-6080(98)00116-6
– ident: ref62
  doi: 10.1016/j.compchemeng.2016.04.026
– ident: ref50
  doi: 10.1137/0728063
– ident: ref21
  doi: 10.1109/IJCNN.1992.227335
– ident: ref23
  doi: 10.1109/TNN.2011.2109737
– ident: ref44
  doi: 10.1137/060674004
– ident: ref42
  doi: 10.1137/0907058
– year: 1975
  ident: ref54
  article-title: A transformation method for constrained-function minimization
– ident: ref13
  doi: 10.1016/j.na.2005.02.015
– ident: ref25
  doi: 10.1109/tac.2022.3222750
– year: 2016
  ident: ref64
  article-title: Forward-mode automatic differentiation in Julia
  publication-title: arXiv:1607.07892
– volume: 6
  start-page: 107
  issue: 2
  year: 1998
  ident: ref11
  article-title: Recurrent neural net learning and vanishing gradient
  publication-title: Int. J. Uncertainity, Fuzziness Knowl.-Based Syst.
  doi: 10.1142/S0218488598000094
– ident: ref60
  doi: 10.1007/978-3-540-89722-4_11
– ident: ref52
  doi: 10.1007/978-3-319-91578-4
– ident: ref46
  doi: 10.2140/pjm.1966.16.1
– ident: ref12
  doi: 10.1109/IJCNN.1999.832591
– ident: ref51
  doi: 10.1137/1.9781611970791
– ident: ref18
  doi: 10.1016/S0925-2312(98)00104-0
– ident: ref28
  doi: 10.1109/tnnls.2021.3109953
– ident: ref2
  doi: 10.1109/tnn.1998.712192
– ident: ref57
  doi: 10.1007/b98874
– start-page: 27
  volume-title: Nonlinear Programming
  year: 1978
  ident: ref41
  article-title: The convergence of variable metric methods for nonlinearly constrained optimization calculations
– volume-title: Proc. ICML
  ident: ref14
  article-title: Learning recurrent neural networks with Hessian-free optimization
– ident: ref24
  doi: 10.1109/TNNLS.2017.2741598
– ident: ref29
  doi: 10.1109/tnnls.2022.3151412
– ident: ref10
  doi: 10.1109/72.279181
– ident: ref45
  doi: 10.1007/s10107-008-0248-3
– ident: ref9
  doi: 10.7551/mitpress/4977.003.0010
– ident: ref66
  doi: 10.21105/joss.00602
– start-page: 1329
  volume-title: Proc. Int. Conf. Mach. Learn.
  ident: ref61
  article-title: Benchmarking deep reinforcement learning for continuous control
– ident: ref49
  doi: 10.1017/S0962492904000212
– year: 2014
  ident: ref55
  article-title: Adam: A method for stochastic optimization
  publication-title: arXiv:1412.6980
– ident: ref34
  doi: 10.1016/j.automatica.2010.06.021
– ident: ref16
  doi: 10.1109/IJCNN.2007.4370969
– year: 2023
  ident: ref68
  article-title: Self-concordant smoothing for convex composite optimization
  publication-title: arXiv:2309.01781
– volume: 3
  volume-title: Proc. Adv. Neural Inf. Process. Syst.
  ident: ref7
  article-title: Reinforcement learning in Markovian and non- Markovian environments
– ident: ref38
  doi: 10.1090/S0025-5718-1970-0274029-X
– ident: ref35
  doi: 10.1093/imamat/6.1.76
– ident: ref36
  doi: 10.1093/comjnl/13.3.317
– ident: ref43
  doi: 10.1137/0719025
– volume: 31
  volume-title: Proc. Adv. Neural Inf. Process. Syst.
  ident: ref4
  article-title: Recurrent world models facilitate policy evolution
– ident: ref39
  doi: 10.1007/BFb0067703
– volume-title: Neural Networks and Learning Machines
  year: 2009
  ident: ref1
– ident: ref19
  doi: 10.1016/j.automatica.2023.111183
– ident: ref33
  doi: 10.1007/s10589-023-00502-2
– year: 2008
  ident: ref59
  article-title: Reinforcement learning with recurrent neural networks
– ident: ref3
  doi: 10.1038/nature14539
– ident: ref53
  doi: 10.1109/ICCV.2015.123
SSID ssj0000605649
Score 2.508924
Snippet This article considers the two-stage approach to solving a partially observable Markov decision process (POMDP): the identification stage and the (optimal)...
SourceID proquest
pubmed
crossref
ieee
SourceType Aggregation Database
Index Database
Enrichment Source
Publisher
StartPage 2762
SubjectTerms Gauss–Newton methods
markov decision processes
Neural networks
numerical optimization
Optimization
Prediction algorithms
Process control
Quadratic programming
Recurrent neural networks
recurrent neural networks (RNNs)
reinforcement learning (RL)
sequential quadratic programming (SQP)
Training
Title An Inexact Sequential Quadratic Programming Method for Learning and Control of Recurrent Neural Networks
URI https://ieeexplore.ieee.org/document/10418043
https://www.ncbi.nlm.nih.gov/pubmed/38294918
https://www.proquest.com/docview/2921117034
Volume 36
WOSCitedRecordID wos001157917100001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE Electronic Library (IEL)
  customDbUrl:
  eissn: 2162-2388
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000605649
  issn: 2162-237X
  databaseCode: RIE
  dateStart: 20120101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS-UwEB5URLz42_XtrhLBm1TTJE2To8iKghZ_wruVNJ2o4PaJ7z3ZP3-TtH14UfDWQxJKv5lmJpn5PoADNIxXpnZJUKxLRGbTxATST5thJW2aG52ZKDaRF4UaDvV116wee2EQMRaf4VF4jHf59chOw1GZ93CRKir4PMznuWybtWYHKtQH5jKGuyyVLGE8H_ZNMlQf3xfF5Z1PB5k44jwLjCjLsMQV00IHvY8Pe1IUWfk83oz7ztnqN994DVa6AJOctBaxDnPYbMBqL95AOl_ehKeThlw0-M_YCbmL9dTe11_IzdTUwSgsuW4rt_76vY1cRZ1p4gNc0hGyPhLT1OS0LXQnI0duw8F9oHoige_Dr1S0BebjLXg4-3N_ep50sguJ5VJNEuYqy6ijtbUyVXWuJRNWInecKl1ZTp0WJjT0OpVh5mruMxjvvBXz0YPNVc23YaEZNbgDxFBEyapcaPSRofALc3TGItOprmTmBpD2H760HSd5kMZ4KWNuQnUZcSsDbmWH2wAOZ3NeW0aOL0dvBVQ-jGwBGcB-D3DpHSrckpgGR9NxybTPiVP_IxQD-NEiP5vdG8zPT1b9Bcss6APHqu7fsDB5m-IuLNr3yfP4bc9b7VDtRav9D4c55uI
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NT9wwEB1RQO1eoBQKC7S4Ercq4NiOEx8RAoG6RFAWaW-R44zbSjSL9gPx87GdZMUFJG452FaUmYnf2DPvARyiZrzUlY28Yl0kEhNH2pN-mgRLaeJUq0QHsYk0z7PRSF23zeqhFwYRQ_EZHvnHcJdfjc3cH5W5CBdxRgX_ACuJEIw27VqLIxXqoLkMgJfFkkWMp6OuTYaq42GeD25dQsjEEeeJ50TpwUeeMSWUV_x4sSsFmZXXEWfYec7X3_nOn2GthZjkpPGJDVjC-gusd_INpI3mTfh7UpPLGp-0mZHbUFHtov2e3Mx15d3CkOumduu_293IVVCaJg7ikpaS9Q_RdUVOm1J3Mrbktz-692RPxDN-uJXypsR8ugV352fD04uoFV6IDJfZLGK2NIxaWhkj46xKlWTCSOSW00yVhlOrhPYtvTZLMLEVdzmMC9-SOfxg0qziX2G5Hte4A0RTRMnKVCh02FC4hTlabZCpWJUysX2Iuw9fmJaV3Itj3BchO6GqCHYrvN2K1m59-LmY89Bwcrw5estb5cXIxiB9-NEZuHAh5e9JdI3j-bRgymXFsfsVij5sN5ZfzO4cZveVVQ_g08XwalAMLvNfe9BjXi041Hjvw_JsMsdvsGoeZ_-mk-_Bd58BPxnpQQ
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=An+Inexact+Sequential+Quadratic+Programming+Method+for+Learning+and+Control+of+Recurrent+Neural+Networks&rft.jtitle=IEEE+transaction+on+neural+networks+and+learning+systems&rft.au=Adeoye%2C+Adeyemi+D&rft.au=Bemporad%2C+Alberto&rft.date=2025-02-01&rft.issn=2162-2388&rft.eissn=2162-2388&rft.volume=36&rft.issue=2&rft.spage=2762&rft_id=info:doi/10.1109%2FTNNLS.2024.3354855&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2162-237X&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2162-237X&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2162-237X&client=summon