{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,20]],"date-time":"2026-08-20T15:33:04Z","timestamp":1787239984102,"version":"build-2736575974"},"reference-count":60,"publisher":"Society for Industrial & Applied Mathematics (SIAM)","issue":"3","funder":[{"DOI":"10.13039\/501100003329","name":"Ministerio de Economia y Competitividad","doi-asserted-by":"crossref","award":["PID2020-112617GB-C22"],"award-info":[{"award-number":["PID2020-112617GB-C22"]}],"id":[{"id":"10.13039\/501100003329","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100003329","name":"Ministerio de Economia y Competitividad","doi-asserted-by":"crossref","award":["TED2021-131390B-I00"],"award-info":[{"award-number":["TED2021-131390B-I00"]}],"id":[{"id":"10.13039\/501100003329","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000921","name":"European Cooperation in Science and Technology","doi-asserted-by":"crossref","award":["MAT-DYN-NET"],"award-info":[{"award-number":["MAT-DYN-NET"]}],"id":[{"id":"10.13039\/501100000921","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Universidad Autonoma de Madrid"},{"DOI":"10.13039\/100005156","name":"Alexander von Humboldt-Stiftung","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100005156","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100019180","name":"HORIZON EUROPE European Research Council","doi-asserted-by":"publisher","award":["HORIZON-MSCA-2021-DN-01"],"award-info":[{"award-number":["HORIZON-MSCA-2021-DN-01"]}],"id":[{"id":"10.13039\/100019180","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"Engineering and Physical Sciences Research Council","doi-asserted-by":"publisher","award":["EP\/T024429\/1"],"award-info":[{"award-number":["EP\/T024429\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100010663","name":"H2020 European Research Council","doi-asserted-by":"publisher","award":["694126-DyCon"],"award-info":[{"award-number":["694126-DyCon"]}],"id":[{"id":"10.13039\/100010663","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["SIAM Rev."],"published-print":{"date-parts":[[2023,8]]},"abstract":"<jats:p>We analyze neural ordinary differential equations (NODEs) from a control theoretical perspective to address some of the main properties and paradigms of deep learning (DL), in particular, data classification and universal approximation. These objectives are tackled and achieved from the perspective of the simultaneous control of systems of NODEs. For instance, in the context of classification, each item to be classified corresponds to a different initial datum for the control problem of the NODE, to be classified, all of them by the same common control, to the location (a subdomain of the Euclidean space) associated to each label. Our proofs are genuinely nonlinear and constructive, allowing us to estimate the complexity of the control strategies we develop. The nonlinear nature of the activation functions governing the dynamics of NODEs under consideration plays a key role in our proofs, since it allows deforming half of the phase space while the other half remains invariant, a property that classical models in mechanics do not fulfill. This very property allows us to build elementary controls inducing specific dynamics and transformations whose concatenation, along with properly chosen hyperplanes, allows us to achieve our goals in finitely many steps. The nonlinearity of the dynamics is assumed to be Lipschitz. Therefore, our results apply also in the particular case of the ReLU activation function. We also present the counterparts in the context of the control of neural transport equations, establishing a link between optimal transport and deep neural networks.<\/jats:p>","DOI":"10.1137\/21m1411433","type":"journal-article","created":{"date-parts":[[2023,8,8]],"date-time":"2023-08-08T04:42:43Z","timestamp":1691469763000},"page":"735-773","source":"Crossref","is-referenced-by-count":57,"title":["Neural ODE Control for Classification, Approximation, and Transport"],"prefix":"10.1137","volume":"65","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4204-459X","authenticated-orcid":true,"given":"Dom\u00e8nec","family":"Ruiz-Balet","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Enrique","family":"Zuazua","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"351","published-online":{"date-parts":[[2023,8,8]]},"reference":[{"key":"atypb1","doi-asserted-by":"publisher","DOI":"10.1007\/s10883-021-09561-2"},{"key":"atypb2","doi-asserted-by":"crossref","unstructured":"J.D. Benamou and Y. Brenier,\n                      The Optimal Time-Continuous Mass Transport Problem and Its Augmented Lagrangian Numerical Resolution\n                      , Report 3356, INRIA, 1998.","DOI":"10.1090\/conm\/226\/03232"},{"key":"atypb3","doi-asserted-by":"publisher","DOI":"10.1007\/s002110050002"},{"key":"atypb4","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-40994-3_25"},{"key":"atypb5","volume-title":"Fractals in Probability and Analysis","author":"Bishop C.","year":"2017"},{"key":"atypb6","doi-asserted-by":"publisher","DOI":"10.1006\/jath.1999.3413"},{"key":"atypb7","first-page":"499","volume":"2","author":"Bousquet O.","year":"2002","journal-title":"J. Mach. Learn. Res."},{"key":"atypb8","doi-asserted-by":"publisher","DOI":"10.1093\/imaiai\/iaac029"},{"key":"atypb9","doi-asserted-by":"publisher","DOI":"10.1006\/jath.2001.3613"},{"key":"atypb10","doi-asserted-by":"publisher","DOI":"10.1023\/B:JMIV.0000011321.19549.88"},{"key":"atypb11","doi-asserted-by":"publisher","DOI":"10.1007\/s002110050258"},{"key":"atypb12","first-page":"6571","author":"Chen R.","year":"2018","journal-title":"Advances in Neural Information Processing Systems"},{"key":"atypb13","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2017.2767707"},{"key":"atypb14","volume-title":"Implicit Bias of Gradient Descent for Wide Two-Layer Neural Networks Trained with the Logistic Loss, preprint, https:\/\/arxiv.org\/abs\/2002.04486","author":"Chizat L.","year":"2020"},{"key":"atypb15","doi-asserted-by":"publisher","DOI":"10.1137\/19M1284117"},{"key":"atypb16","doi-asserted-by":"publisher","DOI":"10.1007\/BF02551274"},{"key":"atypb17","doi-asserted-by":"publisher","DOI":"10.1007\/s00365-021-09548-z"},{"key":"atypb18","doi-asserted-by":"publisher","DOI":"10.1137\/17M1152917"},{"key":"atypb19","doi-asserted-by":"publisher","DOI":"10.1016\/j.jde.2019.11.098"},{"key":"atypb20","doi-asserted-by":"publisher","DOI":"10.1007\/s40304-017-0103-z"},{"key":"atypb21","volume-title":"Large-Time Asymptotics in Deep Learning, preprint, https:\/\/arxiv.org\/abs\/2008.02491","author":"Esteve C.","year":"2020"},{"key":"atypb22","doi-asserted-by":"publisher","DOI":"10.1016\/j.sysconle.2022.105452"},{"key":"atypb23","volume-title":"Fractal Geometry: Mathematical Foundations and Applications","author":"Falconer K.","year":"2004"},{"key":"atypb24","doi-asserted-by":"publisher","DOI":"10.1090\/S0002-9939-1989-0969315-8"},{"key":"atypb25","doi-asserted-by":"publisher","DOI":"10.1017\/S0962492922000046"},{"key":"atypb26","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-0348-5858-8_7"},{"key":"atypb27","doi-asserted-by":"publisher","DOI":"10.1142\/S0219530519410021"},{"key":"atypb28","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6420\/aa9a90"},{"key":"atypb29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"atypb30","doi-asserted-by":"publisher","DOI":"10.1016\/0893-6080(89)90020-8"},{"key":"atypb31","doi-asserted-by":"publisher","DOI":"10.1090\/memo\/0587"},{"key":"atypb32","volume-title":"Conference in Modern Analysis and Probability","author":"Johnson W. B.","year":"1982"},{"key":"atypb33","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2020.2992934"},{"key":"atypb34","doi-asserted-by":"publisher","DOI":"10.1109\/FOCS.2017.64"},{"key":"atypb35","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"atypb36","doi-asserted-by":"publisher","DOI":"10.1016\/S0893-6080(05)80131-5"},{"key":"atypb37","doi-asserted-by":"publisher","DOI":"10.4171\/JEMS\/1221"},{"key":"atypb38","doi-asserted-by":"publisher","DOI":"10.1137\/1030001"},{"key":"atypb39","doi-asserted-by":"publisher","DOI":"10.5802\/afst.1511"},{"key":"atypb40","volume-title":"International Conference on Learning Representations","author":"Madry A.","year":"2018"},{"key":"atypb41","doi-asserted-by":"publisher","DOI":"10.1007\/s10444-004-7634-z"},{"key":"atypb42","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298640"},{"key":"atypb43","doi-asserted-by":"publisher","DOI":"10.1561\/2200000073"},{"key":"atypb44","doi-asserted-by":"publisher","DOI":"10.1137\/140996501"},{"key":"atypb45","doi-asserted-by":"publisher","DOI":"10.1017\/S0962492900002919"},{"key":"atypb46","doi-asserted-by":"publisher","DOI":"10.1201\/9780203749319"},{"key":"atypb47","doi-asserted-by":"publisher","DOI":"10.1007\/s00205-014-0823-0"},{"key":"atypb48","doi-asserted-by":"publisher","DOI":"10.4171\/JEMS\/608"},{"key":"atypb49","doi-asserted-by":"publisher","DOI":"10.1109\/ICIP.1994.413269"},{"key":"atypb50","volume-title":"Interpolation and Approximation via Momentum ResNets and Neural ODEs, preprint, https:\/\/arxiv.org\/abs\/2110.08761","author":"Ruiz-Balet D.","year":"2021"},{"key":"atypb51","first-page":"9276","volume-title":"Proceedings of the 38th International Conference on Machine Learning, PMLR","author":"Sander M.","year":"2021"},{"key":"atypb52","doi-asserted-by":"publisher","DOI":"10.1007\/BF02551238"},{"key":"atypb53","doi-asserted-by":"publisher","DOI":"10.1016\/S0167-6911(97)00002-9"},{"key":"atypb54","doi-asserted-by":"publisher","DOI":"10.1090\/S0002-9939-1992-1094507-X"},{"key":"atypb55","doi-asserted-by":"publisher","DOI":"10.1137\/0317045"},{"key":"atypb56","volume-title":"Universal Approximation Power of Deep Neural Networks via Nonlinear Control Theory, preprint, https:\/\/arxiv.org\/abs\/2007.06007","author":"Tabuada P.","year":"2020"},{"key":"atypb57","volume-title":"Universal Approximation Property of Neural Ordinary Differential Equations, preprint, https:\/\/arxiv.org\/abs\/2012.02414","author":"Teshima T.","year":"2020"},{"key":"atypb58","doi-asserted-by":"publisher","DOI":"10.1137\/S0363012999352716"},{"key":"atypb59","doi-asserted-by":"publisher","DOI":"10.1090\/gsm\/058"},{"key":"atypb60","volume-title":"Grundlehren Math. Wiss. 338","author":"Villani C.","year":"2008"}],"container-title":["SIAM Review"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/epubs.siam.org\/doi\/pdf\/10.1137\/21M1411433","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,20]],"date-time":"2026-08-20T15:10:02Z","timestamp":1787238602000},"score":1,"resource":{"primary":{"URL":"https:\/\/epubs.siam.org\/doi\/10.1137\/21M1411433"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,8]]},"references-count":60,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2023,8]]}},"alternative-id":["10.1137\/21M1411433"],"URL":"https:\/\/doi.org\/10.1137\/21m1411433","relation":{},"ISSN":["0036-1445","1095-7200"],"issn-type":[{"value":"0036-1445","type":"print"},{"value":"1095-7200","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,8]]}}}