[{"ec_funded":1,"publication":"7th Symposium on Foundations of Responsible Computing","publisher":"Schloss Dagstuhl - Leibniz-Zentrum für Informatik","OA_type":"gold","intvolume":"       368","ddc":["000"],"citation":{"short":"N. Kalinin, J.D. Andersson, in:, 7th Symposium on Foundations of Responsible Computing, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026.","mla":"Kalinin, Nikita, and Joel D. Andersson. “Learning Rate Scheduling with Matrix Factorization for Private Training.” <i>7th Symposium on Foundations of Responsible Computing</i>, vol. 368, 2:1-2:21, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026, doi:<a href=\"https://doi.org/10.4230/LIPIcs.FORC.2026.2\">10.4230/LIPIcs.FORC.2026.2</a>.","ieee":"N. Kalinin and J. D. Andersson, “Learning rate scheduling with matrix factorization for private training,” in <i>7th Symposium on Foundations of Responsible Computing</i>, Cambridge, MA; United States, 2026, vol. 368.","ama":"Kalinin N, Andersson JD. Learning rate scheduling with matrix factorization for private training. In: <i>7th Symposium on Foundations of Responsible Computing</i>. Vol 368. Schloss Dagstuhl - Leibniz-Zentrum für Informatik; 2026. doi:<a href=\"https://doi.org/10.4230/LIPIcs.FORC.2026.2\">10.4230/LIPIcs.FORC.2026.2</a>","ista":"Kalinin N, Andersson JD. 2026. Learning rate scheduling with matrix factorization for private training. 7th Symposium on Foundations of Responsible Computing. FORC: Symposium on Foundations of Responsible Computing, LIPIcs, vol. 368, 2:1-2:21.","chicago":"Kalinin, Nikita, and Joel D Andersson. “Learning Rate Scheduling with Matrix Factorization for Private Training.” In <i>7th Symposium on Foundations of Responsible Computing</i>, Vol. 368. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026. <a href=\"https://doi.org/10.4230/LIPIcs.FORC.2026.2\">https://doi.org/10.4230/LIPIcs.FORC.2026.2</a>.","apa":"Kalinin, N., &#38; Andersson, J. D. (2026). Learning rate scheduling with matrix factorization for private training. In <i>7th Symposium on Foundations of Responsible Computing</i> (Vol. 368). Cambridge, MA; United States: Schloss Dagstuhl - Leibniz-Zentrum für Informatik. <a href=\"https://doi.org/10.4230/LIPIcs.FORC.2026.2\">https://doi.org/10.4230/LIPIcs.FORC.2026.2</a>"},"date_created":"2026-06-28T22:01:34Z","publication_identifier":{"isbn":["9783959774192"],"eissn":["1868-8969"]},"volume":368,"date_updated":"2026-06-29T06:56:34Z","oa":1,"oa_version":"Published Version","supplementarymaterial":"no","researchdata_availability":"no","alternative_title":["LIPIcs"],"acknowledgement":"We thank Rasmus Pagh, Christoph Lampert and Jalaj Upadhyay for valuable\r\ncomments on an early draft. We thank Ryan Mckenna for a fruitful discussion on the experiment\r\ndesign. We thank Antti Honkela for sharing insights on learning rate scheduling and DP.\r\nNikita P. Kalinin: Funded in part by the Austrian Science Fund (FWF) [10.55776/COE12].\r\nJoel Daniel Andersson: Funded by the European Union. Views and opinions expressed are however\r\nthose of the author(s) only and do not necessarily reflect those of the European Union or the European\r\nResearch Council Executive Agency. Neither the European Union nor the granting authority can be\r\nheld responsible for them. This project has received funding from the European Research Council\r\n(ERC) under the European Union’s Horizon 2020 research and innovation programme (MoDynStruct,\r\nNo. 101019564). Additional funding by Providentia, a Data Science Distinguished Investigator grant\r\nfrom Novo Nordisk Fonden, with additional support from VILLUM Investigator grant 54451.\r\n","status":"public","keyword":["differential privacy","machine learning","matrix factorization"],"file_date_updated":"2026-06-29T06:55:23Z","scopus_import":"1","conference":{"location":"Cambridge, MA; United States","end_date":"2026-06-05","start_date":"2026-06-03","name":"FORC: Symposium on Foundations of Responsible Computing"},"_id":"22146","OA_place":"publisher","has_accepted_license":"1","abstract":[{"lang":"eng","text":"We study differentially private model training with stochastic gradient descent under learning rate scheduling and correlated noise. Although correlated noise, in particular via matrix factorizations, has been shown to improve accuracy, prior theoretical work focused primarily on the prefix-sum workload. That workload assumes a constant learning rate, whereas in practice learning rate schedules are widely used to accelerate training and improve convergence. We close this gap by deriving general upper and lower bounds for a broad class of learning rate schedules in both single- and multi-epoch settings. Building on these results, we propose a learning-rate-aware factorization that achieves improvements over prefix-sum factorizations under both MaxSE and MeanSE error metrics. Our theoretical analysis yields memory-efficient constructions suitable for practical deployment, and experiments on CIFAR-10 and IMDB datasets confirm that schedule-aware factorizations improve accuracy in private training."}],"type":"conference","article_number":"2:1-2:21","language":[{"iso":"eng"}],"external_id":{"arxiv":["2511.17994"]},"month":"06","day":"01","author":[{"id":"4b14526e-14d2-11ed-ba64-c14c9553d137","last_name":"Kalinin","full_name":"Kalinin, Nikita","first_name":"Nikita"},{"first_name":"Joel D","full_name":"Andersson, Joel D","last_name":"Andersson","id":"4a893819-d954-11f0-89b1-e360bad9ccc5"}],"das_tickbox":"0","date_published":"2026-06-01T00:00:00Z","tmp":{"image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","short":"CC BY (4.0)","legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode"},"article_processing_charge":"No","arxiv":1,"corr_author":"1","publication_status":"published","year":"2026","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","department":[{"_id":"ChLa"},{"_id":"GradSch"},{"_id":"MoHe"}],"title":"Learning rate scheduling with matrix factorization for private training","file":[{"relation":"main_file","date_created":"2026-06-29T06:55:23Z","content_type":"application/pdf","checksum":"c661f016d3861a1c1b590b87a744d087","success":1,"file_size":1231914,"date_updated":"2026-06-29T06:55:23Z","creator":"dernst","file_name":"2026_LIPIcsFORC_Kalinin.pdf","file_id":"22149","access_level":"open_access"}],"doi":"10.4230/LIPIcs.FORC.2026.2","project":[{"name":"The design and evaluation of modern fully dynamic data structures","call_identifier":"H2020","_id":"bd9ca328-d553-11ed-ba76-dc4f890cfe62","grant_number":"101019564"}],"quality_controlled":"1"}]