{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T06:47:49Z","timestamp":1790232469559,"version":"4.1.0"},"reference-count":34,"publisher":"Society for Industrial & Applied Mathematics (SIAM)","issue":"2","funder":[{"DOI":"10.13039\/100004330","name":"GlaxoSmithKline","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100004330","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100017899","name":"Masason Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100017899","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"publisher","award":["EP\/V029428\/1"],"award-info":[{"award-number":["EP\/V029428\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"publisher","award":["EP\/S026045\/1"],"award-info":[{"award-number":["EP\/S026045\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"publisher","award":["EP\/T003553\/1"],"award-info":[{"award-number":["EP\/T003553\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"publisher","award":["EP\/N014588\/1"],"award-info":[{"award-number":["EP\/N014588\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"publisher","award":["EP\/T017961\/1"],"award-info":[{"award-number":["EP\/T017961\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Wellcome Innovator","award":["215733\/Z\/19\/Z"],"award-info":[{"award-number":["215733\/Z\/19\/Z"]}]},{"name":"Wellcome Innovator","award":["221633\/Z\/20\/Z"],"award-info":[{"award-number":["221633\/Z\/20\/Z"]}]},{"name":"European Union Horizon 2020 Research and Innovation Programme","award":["777826 NoMADS"],"award-info":[{"award-number":["777826 NoMADS"]}]},{"name":"Cantab Capital Institute for the Mathematics of Information"},{"DOI":"10.13039\/100012338","name":"Alan Turing Institute","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100012338","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["SIAM Journal on Mathematics of Data Science"],"published-print":{"date-parts":[[2023,6,30]]},"abstract":"<jats:p>Abstract.<\/jats:p>\n                  <jats:p>Learning-to-optimize is an emerging framework that seeks to speed up the solution of certain optimization problems by leveraging training data. Learned optimization solvers have been shown to outperform classical optimization algorithms in terms of convergence speed, especially for convex problems. Many existing data-driven optimization methods are based on parameterizing the update step and learning the optimal parameters (typically scalars) from the available data. We propose a novel functional parameterization approach for learned convex optimization solvers based on the classical mirror descent (MD) algorithm. Specifically, we seek to learn the optimal Bregman distance in MD by modeling the underlying convex function using an input-convex neural network (ICNN). The parameters of the ICNN are learned by minimizing the target objective function evaluated at the MD iterate after a predetermined number of iterations. The inverse of the mirror map is modeled approximately using another neural network, as the exact inverse is intractable to compute. We derive convergence rate bounds for the proposed learned mirror descent approach with an approximate inverse mirror map and perform extensive numerical evaluation on various convex problems such as image inpainting, denoising, and learning a two-class support vector machine classifier and a multiclass linear classifier on fixed features.<\/jats:p>","DOI":"10.1137\/22m1508613","type":"journal-article","created":{"date-parts":[[2023,6,26]],"date-time":"2023-06-26T11:07:26Z","timestamp":1687777646000},"page":"558-587","source":"Crossref","is-referenced-by-count":5,"title":["Data-Driven Mirror Descent with Input-Convex Neural Networks"],"prefix":"10.1137","volume":"5","author":[{"given":"Hong Ye","family":"Tan","sequence":"first","affiliation":[{"name":"Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, CB3 0WA, UK."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Subhadip","family":"Mukherjee","sequence":"additional","affiliation":[{"name":"Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, CB3 0WA, UK."},{"name":"Department of Electronics and Electrical Communication Engineering, Indian Institute of Technology, Kharagpur, India."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Junqi","family":"Tang","sequence":"additional","affiliation":[{"name":"Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, CB3 0WA, UK."},{"name":"School of Mathematics, University of Birmingham, Birmingham, B15 2TT, UK."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Carola-Bibiane","family":"Sch\u00f6nlieb","sequence":"additional","affiliation":[{"name":"Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, CB3 0WA, UK."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"351","published-online":{"date-parts":[[2023,6,26]]},"reference":[{"key":"ref1","unstructured":"B. Amos , \nL. Xu , and \nJ. Z. Kolter , Input convex neural networks, in Proceedings of the 34th International Conference on Machine Learning, Proc. Mach. Learn. Res. 70, PMLR, 2017, pp. 146\u2013155."},{"key":"ref2","unstructured":"M. Andrychowicz , \nM. Denil , \nS. G\u00f3mez , \nM. W. Hoffman , \nD. Pfau , \nT. Schaul , \nB. Shillingford , and \nN. de Freitas , Learning to learn by gradient descent by gradient descent, in NIPS\u201916: Proceedings of the 30th International Conference on Neural Information Processing Systems, \nD. Lee , \nM. Sugiyama , \nU. Luxburg , \nI. Guyon , and \nR. Garnett , eds. Curran Associates, 2016, pp. 3988\u20133996."},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1137\/18M1207685"},{"key":"ref4","unstructured":"S. Banert , \nJ. Rudzusika , \nO. \u00d6ktem , and \nJ. Adler , Accelerated Forward-Backward Optimization Using Deep Learning, https:\/\/arxiv.org\/abs\/2105.05210, 2021."},{"key":"ref5","doi-asserted-by":"crossref","first-page":"167","DOI":"10.1016\/S0167-6377(02)00231-6","volume":"31","author":"Beck A.","year":"2003","journal-title":"Oper. Res. Lett."},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1137\/S1052623499354564"},{"key":"ref7","volume-title":"Pattern Recognition and Machine Learning (Information Science and Statistics)","author":"Bishop C. M.","year":"2006"},{"key":"ref8","first-page":"1","volume-title":"2019 International Joint Conference on Neural Networks (IJCNN)","author":"Bock S.","year":"2019"},{"key":"ref9","doi-asserted-by":"crossref","first-page":"231","DOI":"10.1561\/2200000050","volume":"8","author":"Bubeck S.","year":"2015","journal-title":"Found. Trends Mach. Learn."},{"key":"ref10","doi-asserted-by":"crossref","first-page":"120","DOI":"10.1007\/s10851-010-0251-1","volume":"40","author":"Chambolle A.","year":"2010","journal-title":"J. Math. Imaging Vis."},{"key":"ref11","unstructured":"S. H. Chan , \nX. Wang , and \nO. A. Elgendy , Plug-and-Play ADMM for Image Restoration: Fixed Point Convergence and Applications, CoRR, http:\/\/arxiv.org\/abs\/1605.01710, 2016."},{"key":"ref12","unstructured":"A. Coates , \nA. Ng , and \nH. Lee , An analysis of single-layer networks in unsupervised feature learning, in Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, \nG. Gordon , \nD. Dunson , and \nM. Dud\u00edk , eds.Proc. Mach. Learn. Res. 15, PMLR, Fort Lauderdale, FL, 2011, pp. 215\u2013223."},{"key":"ref13","unstructured":"K. Gregor  and \nY. LeCun , Learning fast approximations of sparse coding, in Proceedings of the 27th International Conference on International Conference on Machine Learning, ICML\u201910, Omnipress, 2010, pp. 399\u2013406."},{"key":"ref14","series-title":"Proc. Mach. Learn. Res. 130","first-page":"2305","volume-title":"Proceedings of the 24th International Conference on Artificial Intelligence and Statistics","author":"Gunasekar S.","year":"2021"},{"key":"ref15","doi-asserted-by":"crossref","first-page":"405","DOI":"10.1007\/s10589-021-00273-8","volume":"79","author":"Hanzely F.","year":"2021","journal-title":"Comput. Optim. Appl."},{"key":"ref16","first-page":"2845","volume":"28","author":"Krichene W.","year":"2015","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref17","doi-asserted-by":"crossref","first-page":"365","DOI":"10.1007\/s10107-010-0434-y","volume":"133","author":"Lan G.","year":"2012","journal-title":"Math. Program."},{"key":"ref18","doi-asserted-by":"crossref","first-page":"167","DOI":"10.1007\/s10107-017-1173-0","volume":"171","author":"Lan G.","year":"2018","journal-title":"Math. Program."},{"key":"ref19","unstructured":"K. Li  and \nJ. Malik , Learning to Optimize, CoRR, https:\/\/arxiv.org\/abs\/1606.01885, 2016."},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1137\/16M1099546"},{"key":"ref21","unstructured":"N. Maheswaranathan , \nD. Sussillo , \nL. Metz , \nR. Sun , and \nJ. Sohl-Dickstein , Reverse Engineering Learned Optimizers Reveals Known and Novel Mechanisms, CoRR, https:\/\/arxiv.org\/abs\/2011.02159, 2020."},{"key":"ref22","author":"Mukherjee S.","year":"2020","journal-title":"Learned convex regularizers for inverse problems"},{"key":"ref23","series-title":"Wiley-Interscience Ser. Discrete Math.","volume-title":"Problem Complexity and Method Efficiency in Optimization","author":"Nemirovsky A. S.","year":"1983"},{"key":"ref24","doi-asserted-by":"crossref","first-page":"125","DOI":"10.1007\/s10107-012-0629-5","volume":"140","author":"Nesterov Y.","year":"2013","journal-title":"Math. Program."},{"key":"ref25","doi-asserted-by":"crossref","first-page":"411","DOI":"10.1007\/s10994-014-5474-8","volume":"99","author":"Orabona F.","year":"2015","journal-title":"Mach. Learn."},{"key":"ref26","first-page":"8024","volume-title":"Advances in Neural Information Processing Systems 32","author":"Paszke A.","year":"2019"},{"key":"ref27","unstructured":"S. J. Reddi , \nS. Kale , and \nS. Kumar , On the Convergence of Adam and Beyond, CoRR, http:\/\/arxiv.org\/abs\/1904.09237, 2019."},{"key":"ref28","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-642-02431-3","volume-title":"Variational Analysis","author":"Rockafellar R.","year":"1998"},{"key":"ref29","unstructured":"E. K. Ryu , \nJ. Liu , \nS. Wang , \nX. Chen , \nZ. Wang , and \nW. Yin , Plug-and-Play Methods Provably Converge with Properly Trained Denoisers, CoRR, http:\/\/arxiv.org\/abs\/1905.05406, 2019"},{"key":"ref30","unstructured":"P. Tseng , On Accelerated Proximal Gradient Methods for Convex-Concave Optimization, Technical report, University of Washington, Seattle, 2008."},{"key":"ref31","volume-title":"Meta-learning: A Survey","author":"Vanschoren J.","year":"2018"},{"key":"ref32","doi-asserted-by":"crossref","first-page":"945","DOI":"10.1109\/GlobalSIP.2013.6737048","volume-title":"2013 IEEE Global Conference on Signal and Information Processing","author":"Venkatakrishnan S. V.","year":"2013"},{"key":"ref33","unstructured":"P. Xu , \nT. Wang , and \nQ. Gu , Accelerated stochastic mirror descent: From continuous-time dynamics to discrete-time algorithms, in Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR, 2018, pp. 1087\u20131096."},{"key":"ref34","unstructured":"F. Zou , \nL. Shen , \nZ. Jie , \nW. Zhang , and \nW. Liu , A Sufficient Condition for Convergences of Adam and RMSProp, CoRR, https:\/\/arxiv.org\/abs\/1811.09358, 2018."}],"container-title":["SIAM Journal on Mathematics of Data Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/epubs.siam.org\/doi\/pdf\/10.1137\/22M1508613","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T15:15:14Z","timestamp":1787325314000},"score":1,"resource":{"primary":{"URL":"https:\/\/epubs.siam.org\/doi\/10.1137\/22M1508613"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,26]]},"references-count":34,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2023,6,30]]}},"alternative-id":["10.1137\/22M1508613"],"URL":"https:\/\/doi.org\/10.1137\/22m1508613","relation":{},"ISSN":["2577-0187"],"issn-type":[{"value":"2577-0187","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,6,26]]}}}