[{"alternative_title":["LIPIcs"],"department":[{"_id":"ChLa"},{"_id":"GradSch"},{"_id":"MoHe"}],"_id":"22146","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","external_id":{"arxiv":["2511.17994"]},"conference":{"location":"Cambridge, MA; United States","start_date":"2026-06-03","end_date":"2026-06-05","name":"FORC: Symposium on Foundations of Responsible Computing"},"publisher":"Schloss Dagstuhl - Leibniz-Zentrum für Informatik","month":"06","oa_version":"Published Version","publication_identifier":{"isbn":["9783959774192"],"eissn":["1868-8969"]},"volume":368,"researchdata_availability":"no","intvolume":"       368","article_number":"2:1-2:21","quality_controlled":"1","publication":"7th Symposium on Foundations of Responsible Computing","citation":{"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.","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>","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>.","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.","short":"N. Kalinin, J.D. Andersson, in:, 7th Symposium on Foundations of Responsible Computing, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026."},"year":"2026","date_created":"2026-06-28T22:01:34Z","date_updated":"2026-09-16T07:37:21Z","author":[{"full_name":"Kalinin, Nikita","first_name":"Nikita","id":"4b14526e-14d2-11ed-ba64-c14c9553d137","last_name":"Kalinin"},{"full_name":"Andersson, Joel D","first_name":"Joel D","id":"4a893819-d954-11f0-89b1-e360bad9ccc5","last_name":"Andersson"}],"das_tickbox":"0","date_published":"2026-06-01T00:00:00Z","day":"01","doi":"10.4230/LIPIcs.FORC.2026.2","scopus_import":"1","abstract":[{"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.","lang":"eng"}],"language":[{"iso":"eng"}],"tmp":{"name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","image":"/images/cc_by.png","short":"CC BY (4.0)"},"user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","file_date_updated":"2026-06-29T06:55:23Z","has_accepted_license":"1","article_processing_charge":"No","oa":1,"keyword":["differential privacy","machine learning","matrix factorization"],"corr_author":"1","OA_place":"publisher","license":"https://creativecommons.org/licenses/by/4.0/","supplementarymaterial":"no","type":"conference","OA_type":"gold","ddc":["000"],"ec_funded":1,"status":"public","arxiv":1,"fulldoi":"https://doi.org/10.4230/LIPIcs.FORC.2026.2","publication_status":"published","title":"Learning rate scheduling with matrix factorization for private training","project":[{"call_identifier":"H2020","name":"The design and evaluation of modern fully dynamic data structures","grant_number":"101019564","_id":"bd9ca328-d553-11ed-ba76-dc4f890cfe62"},{"name":"Bilateral Artificial Intelligence (Lampert)","_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","grant_number":"COE12"}],"file":[{"date_updated":"2026-06-29T06:55:23Z","relation":"main_file","success":1,"file_name":"2026_LIPIcsFORC_Kalinin.pdf","access_level":"open_access","date_created":"2026-06-29T06:55:23Z","content_type":"application/pdf","file_size":1231914,"file_id":"22149","checksum":"c661f016d3861a1c1b590b87a744d087","creator":"dernst"}]},{"supervisor":[{"id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","first_name":"Christoph","last_name":"Lampert","full_name":"Lampert, Christoph","orcid":"0000-0001-8622-7887"}],"date_created":"2026-02-09T14:59:53Z","year":"2026","citation":{"ama":"Scott JA. Data heterogeneity and personalization in federated learning. 2026. doi:<a href=\"https://doi.org/10.15479/AT-ISTA-21198\">10.15479/AT-ISTA-21198</a>","ista":"Scott JA. 2026. Data heterogeneity and personalization in federated learning. Institute of Science and Technology Austria.","apa":"Scott, J. A. (2026). <i>Data heterogeneity and personalization in federated learning</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/AT-ISTA-21198\">https://doi.org/10.15479/AT-ISTA-21198</a>","chicago":"Scott, Jonathan A. “Data Heterogeneity and Personalization in Federated Learning.” Institute of Science and Technology Austria, 2026. <a href=\"https://doi.org/10.15479/AT-ISTA-21198\">https://doi.org/10.15479/AT-ISTA-21198</a>.","mla":"Scott, Jonathan A. <i>Data Heterogeneity and Personalization in Federated Learning</i>. Institute of Science and Technology Austria, 2026, doi:<a href=\"https://doi.org/10.15479/AT-ISTA-21198\">10.15479/AT-ISTA-21198</a>.","ieee":"J. A. Scott, “Data heterogeneity and personalization in federated learning,” Institute of Science and Technology Austria, 2026.","short":"J.A. Scott, Data Heterogeneity and Personalization in Federated Learning, Institute of Science and Technology Austria, 2026."},"date_updated":"2026-10-01T09:29:03Z","author":[{"full_name":"Scott, Jonathan A","first_name":"Jonathan A","id":"e499926b-f6e0-11ea-865d-9c63db0031e8","last_name":"Scott"}],"acknowledged_ssus":[{"_id":"ScienComp"}],"doi":"10.15479/AT-ISTA-21198","day":"09","date_published":"2026-02-09T00:00:00Z","abstract":[{"lang":"eng","text":"In recent years there has been a massive increase in the amount of data generated in a\r\ndecentralized manner. Ever more powerful edge devices, such as smartphones, have become\r\nubiquitous in most societies on earth. Through text typed, photos taken and apps used,\r\nthese devices, which we refer to as clients, generate enormous amounts of high quality and\r\ncomplex data. Moreover, the nature of these devices means the data they generate is often\r\nsensitive and privacy concerns prevent it being gathered and stored in a central location. This\r\npresents a challenge to the modern machine learning paradigm that requires central access\r\nto large amounts of data. Federated learning (FL) has emerged as one of the answers to\r\nthis problem. Rather than bringing the data to the model, FL sends the model to the data.\r\nModel training takes place on device, with periodically synchronized updates, allowing data to\r\nremain locally stored. While this approach offers significant privacy advantages it comes with\r\nits own set of unique challenges. These include: data heterogeneity, the notion that different\r\ndevices generate data in distinct ways which can negatively impact training dynamics; systems\r\nheterogeneity, meaning that different devices may have differing hardware specifications; high\r\ncommunication costs, which are induced by the repeated transferring of models over the\r\nnetwork and low device computational power, which limits the use of larger models on device.\r\nIn this thesis we present a range of methods for federated learning. We focus primarily on\r\nthe challenge of data heterogeneity, though the methods presented are designed to be well\r\nadapted to the other challenges of a federated setting, such as the constraints of limited\r\ncompute and communication overhead. We first present a method for explicitly modeling client\r\ndata heterogeneity. The approach formulates clients as samples from a certain probability\r\ndistribution and infers the parameters of this distribution from the available training clients.\r\nThis learned distribution then represents the heterogeneity present among the clients and can\r\nbe sampled from in order to create new simulated clients that are similar to the real clients we\r\nhave observed so far. Following this we present two methods for directly dealing with data\r\nheterogeneity through personalization. Highly heterogeneous client data distributions can mean\r\nthat learning a single global model becomes suboptimal, and some form of personalization of\r\nmodels to each individual client is required. Our approaches are based around hypernetworks,\r\nwhich we use to generate personalized model parameters without the need for additional\r\ntraining or finetuning. In the first approach we focus on generating full parameterizations of\r\nclient models using learned embeddings of client data and labels, with a hypernetwork located\r\non the central server. In the second approach we address the more challenging scenario where\r\nwe want to generate a personalized model for a client without any label information. The\r\nhypernetwork is trained to generate a low dimensional representation of a client’s personalized\r\nmodel parameters, allowing it to be transferred to and run on the client devices. In our final\r\npresented method, we change our focus and rather than aim to directly address the challenge\r\nof data heterogeneity, we instead ensure we are unaffected by it. This is done in the context\r\nof k-means clustering and we present a method for federated clustering with a focus on added\r\nprivacy guarantees."}],"_id":"21198","alternative_title":["ISTA Thesis"],"department":[{"_id":"GradSch"},{"_id":"ChLa"}],"acknowledgement":"This research was funded in part by the Austrian Science Fund (FWF)\r\n[10.55776/COE12]. Furthermore, the candidate acknowledges the support from the Scientific\r\nService Units (SSU) of ISTA through resources provided by Scientific Computing (SciComp).","oa_version":"Published Version","month":"02","publisher":"Institute of Science and Technology Austria","publication_identifier":{"issn":["2663-337X"]},"type":"dissertation","ddc":["005"],"degree_awarded":"PhD","status":"public","fulldoi":"https://doi.org/10.15479/AT-ISTA-21198","doi_confirm":"1","title":"Data heterogeneity and personalization in federated learning","project":[{"name":"Bilateral Artificial Intelligence (Lampert)","_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","grant_number":"COE12"}],"file":[{"relation":"source_file","date_updated":"2026-02-17T11:46:22Z","file_name":"2026_Scott_Jonathan_Thesis_Source.zip","access_level":"closed","date_created":"2026-02-17T11:46:22Z","content_type":"application/zip","file_size":272379252,"file_id":"21298","checksum":"121c1d968bd86f3630aa7e81d5bbbcb0","creator":"jscott"},{"content_type":"application/pdf","date_created":"2026-02-27T10:25:41Z","file_name":"2026_Jonathan_Scott_Thesis.pdf","access_level":"open_access","success":1,"date_updated":"2026-02-27T10:25:41Z","relation":"main_file","creator":"jscott","checksum":"6e3e08ba474bbee8511cc8a839ab2077","file_size":15220298,"file_id":"21366"}],"publication_status":"published","page":"158","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","language":[{"iso":"eng"}],"has_accepted_license":"1","article_processing_charge":"No","file_date_updated":"2026-02-27T10:25:41Z","oa":1,"related_material":{"record":[{"relation":"part_of_dissertation","status":"public","id":"21207"},{"relation":"part_of_dissertation","status":"public","id":"20819"},{"status":"public","id":"17411","relation":"part_of_dissertation"},{"id":"18120","status":"public","relation":"part_of_dissertation"}]},"OA_place":"publisher","corr_author":"1"},{"arxiv":1,"project":[{"grant_number":"101020093","_id":"62781420-2b32-11ec-9570-8d9b63373d4d","name":"Vigilant Algorithmic Monitoring of Software","call_identifier":"H2020"},{"_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","grant_number":"COE12","name":"Bilateral Artificial Intelligence (Lampert)"}],"title":"Logic gate neural networks are good for verification","file":[{"file_id":"20314","file_size":295466,"creator":"dernst","checksum":"90a32defed34787e771a5c1623b6b0d2","relation":"main_file","date_updated":"2025-09-09T08:10:13Z","success":1,"content_type":"application/pdf","date_created":"2025-09-09T08:10:13Z","access_level":"open_access","file_name":"2025_NeuS_Kresse.pdf"}],"publication_status":"published","ddc":["000"],"ec_funded":1,"OA_type":"diamond","type":"conference","status":"public","OA_place":"publisher","corr_author":"1","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","language":[{"iso":"eng"}],"oa":1,"file_date_updated":"2025-09-09T08:10:13Z","article_processing_charge":"No","has_accepted_license":"1","acknowledged_ssus":[{"_id":"ScienComp"}],"abstract":[{"lang":"eng","text":"Learning-based systems are increasingly deployed across various domains, yet the complexity of traditional neural networks poses significant challenges for formal verification. Unlike conventional neural networks, learned Logic Gate Networks (LGNs) replace multiplications with Boolean logic gates, yielding a sparse, netlist-like architecture that is inherently more amenable to symbolic verification, while still delivering promising performance. In this paper, we introduce a SAT encoding for verifying global robustness and fairness in LGNs. We evaluate our method on five benchmark datasets, including a newly constructed 5-class variant, and find that LGNs are both verification-friendly and maintain strong predictive performance."}],"scopus_import":"1","day":"01","date_published":"2025-06-01T00:00:00Z","year":"2025","date_created":"2025-09-07T22:01:34Z","citation":{"short":"F. Kresse, E. Yu, C. Lampert, T.A. Henzinger, in:, 2nd International Conferenceon Neuro-Symbolic Systems, ML Research Press, 2025.","ieee":"F. Kresse, E. Yu, C. Lampert, and T. A. Henzinger, “Logic gate neural networks are good for verification,” in <i>2nd International Conferenceon Neuro-Symbolic Systems</i>, Philadephia, PA, United States, 2025, vol. 288.","mla":"Kresse, Fabian, et al. “Logic Gate Neural Networks Are Good for Verification.” <i>2nd International Conferenceon Neuro-Symbolic Systems</i>, vol. 288, 26, ML Research Press, 2025.","chicago":"Kresse, Fabian, Emily Yu, Christoph Lampert, and Thomas A Henzinger. “Logic Gate Neural Networks Are Good for Verification.” In <i>2nd International Conferenceon Neuro-Symbolic Systems</i>, Vol. 288. ML Research Press, 2025.","apa":"Kresse, F., Yu, E., Lampert, C., &#38; Henzinger, T. A. (2025). Logic gate neural networks are good for verification. In <i>2nd International Conferenceon Neuro-Symbolic Systems</i> (Vol. 288). Philadephia, PA, United States: ML Research Press.","ista":"Kresse F, Yu E, Lampert C, Henzinger TA. 2025. Logic gate neural networks are good for verification. 2nd International Conferenceon Neuro-Symbolic Systems. NeuS: International Conferenceon Neuro-Symbolic Systems, PMLR, vol. 288, 26.","ama":"Kresse F, Yu E, Lampert C, Henzinger TA. Logic gate neural networks are good for verification. In: <i>2nd International Conferenceon Neuro-Symbolic Systems</i>. Vol 288. ML Research Press; 2025."},"publication":"2nd International Conferenceon Neuro-Symbolic Systems","author":[{"full_name":"Kresse, Fabian","last_name":"Kresse","first_name":"Fabian","id":"faff3c84-23f6-11ef-9085-e5187b51c604"},{"orcid":"0000-0002-4993-773X","full_name":"Yu, Zhengqi","last_name":"Yu","first_name":"Zhengqi","id":"20aa2ae8-f2f1-11ed-bbfa-8205053f1342"},{"orcid":"0000-0001-8622-7887","full_name":"Lampert, Christoph","last_name":"Lampert","first_name":"Christoph","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87"},{"orcid":"0000-0002-2985-7724","full_name":"Henzinger, Thomas A","last_name":"Henzinger","id":"40876CD8-F248-11E8-B48F-1D18A9856A87","first_name":"Thomas A"}],"date_updated":"2026-09-16T07:02:18Z","intvolume":"       288","volume":288,"publication_identifier":{"eissn":["2640-3498"]},"quality_controlled":"1","article_number":"26","acknowledgement":"This work is supported in part by the ERC grant under Grant No. ERC-2020-AdG 101020093 and\r\nthe Austrian Science Fund (FWF) [10.55776/COE12]. This research was supported by the Scientific\r\nService Units (SSU) of ISTA through resources provided by Scientific Computing (SciComp).","_id":"20296","alternative_title":["PMLR"],"department":[{"_id":"ChLa"},{"_id":"ToHe"}],"oa_version":"Published Version","month":"06","publisher":"ML Research Press","conference":{"start_date":"2025-05-28","name":"NeuS: International Conferenceon Neuro-Symbolic Systems","end_date":"2025-05-30","location":"Philadephia, PA, United States"},"external_id":{"arxiv":["2505.19932"]}},{"acknowledgement":"This research was funded in part by the Austrian Science Fund (FWF) [10.55776/COE12] and supported by the Scientific Service Units (SSU) of ISTA through resources provided by Scientific Computing (SciComp).\r\n","_id":"20819","department":[{"_id":"ChLa"},{"_id":"MoHe"}],"alternative_title":["PMLR"],"oa_version":"Published Version","month":"05","publisher":"ML Research Press","conference":{"location":"Vancouver, Canada","name":"ICML: International Conference on Machine Learning","end_date":"2025-07-19","start_date":"2025-07-13"},"external_id":{"arxiv":["2506.05408"]},"intvolume":"       267","volume":267,"publication_identifier":{"eissn":["2640-3498"]},"quality_controlled":"1","date_created":"2025-12-14T23:02:05Z","year":"2025","citation":{"ista":"Scott JA, Lampert C, Saulpic D. 2025. Differentially private federated k-means clustering with server-side data. 42nd International Conference on Machine Learning. ICML: International Conference on Machine Learning, PMLR, vol. 267, 53757–53790.","ama":"Scott JA, Lampert C, Saulpic D. Differentially private federated k-means clustering with server-side data. In: <i>42nd International Conference on Machine Learning</i>. Vol 267. ML Research Press; 2025:53757-53790.","apa":"Scott, J. A., Lampert, C., &#38; Saulpic, D. (2025). Differentially private federated k-means clustering with server-side data. In <i>42nd International Conference on Machine Learning</i> (Vol. 267, pp. 53757–53790). Vancouver, Canada: ML Research Press.","chicago":"Scott, Jonathan A, Christoph Lampert, and David Saulpic. “Differentially Private Federated K-Means Clustering with Server-Side Data.” In <i>42nd International Conference on Machine Learning</i>, 267:53757–90. ML Research Press, 2025.","mla":"Scott, Jonathan A., et al. “Differentially Private Federated K-Means Clustering with Server-Side Data.” <i>42nd International Conference on Machine Learning</i>, vol. 267, ML Research Press, 2025, pp. 53757–90.","ieee":"J. A. Scott, C. Lampert, and D. Saulpic, “Differentially private federated k-means clustering with server-side data,” in <i>42nd International Conference on Machine Learning</i>, Vancouver, Canada, 2025, vol. 267, pp. 53757–53790.","short":"J.A. Scott, C. Lampert, D. Saulpic, in:, 42nd International Conference on Machine Learning, ML Research Press, 2025, pp. 53757–53790."},"publication":"42nd International Conference on Machine Learning","author":[{"full_name":"Scott, Jonathan A","first_name":"Jonathan A","id":"e499926b-f6e0-11ea-865d-9c63db0031e8","last_name":"Scott"},{"last_name":"Lampert","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","first_name":"Christoph","orcid":"0000-0001-8622-7887","full_name":"Lampert, Christoph"},{"full_name":"Saulpic, David","last_name":"Saulpic","first_name":"David","id":"f8e48cf0-b0ff-11ed-b0e9-b4c35598f964"}],"date_updated":"2026-09-16T07:26:12Z","acknowledged_ssus":[{"_id":"ScienComp"}],"abstract":[{"text":"Clustering is a cornerstone of data analysis that is particularly suited to identifying coherent subgroups or substructures in unlabeled data, as are generated continuously in large amounts these days. However, in many cases traditional clustering methods are not applicable, because data are increasingly being produced and stored in a distributed way, e.g. on edge devices, and privacy concerns prevent it from being transferred to a central server. To address this challenge, we present FedDP-KMeans, a new algorithm for \r\n-means clustering that is fully-federated as well as differentially private. Our approach leverages (potentially small and out-of-distribution) server-side data to overcome the primary challenge of differentially private clustering methods: the need for a good initialization. Combining our initialization with a simple federated DP-Lloyds algorithm we obtain an algorithm that achieves excellent results on synthetic and real-world benchmark tasks. We also provide a theoretical analysis of our method that provides bounds on the convergence speed and cluster identification success.","lang":"eng"}],"scopus_import":"1","date_published":"2025-05-01T00:00:00Z","day":"01","page":"53757-53790","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","tmp":{"name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","image":"/images/cc_by.png","short":"CC BY (4.0)"},"language":[{"iso":"eng"}],"oa":1,"related_material":{"record":[{"relation":"dissertation_contains","id":"21198","status":"public"}]},"file_date_updated":"2025-12-16T12:38:29Z","has_accepted_license":"1","article_processing_charge":"No","OA_place":"publisher","corr_author":"1","ddc":["000"],"OA_type":"gold","type":"conference","status":"public","arxiv":1,"file":[{"file_id":"20829","file_size":746612,"checksum":"815b32b463023ca21e569c2158745c15","creator":"dernst","success":1,"date_updated":"2025-12-16T12:38:29Z","relation":"main_file","file_name":"2025_ICML_Scott.pdf","access_level":"open_access","date_created":"2025-12-16T12:38:29Z","content_type":"application/pdf"}],"project":[{"name":"Bilateral Artificial Intelligence (Lampert)","_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","grant_number":"COE12"}],"title":"Differentially private federated k-means clustering with server-side data","publication_status":"published"},{"oa":1,"article_processing_charge":"No","language":[{"iso":"eng"}],"page":"64002-64047","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","supplementarymaterial":"no","corr_author":"1","OA_place":"publisher","status":"public","type":"conference","OA_type":"free access","publication_status":"published","project":[{"grant_number":"COE12","_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","name":"Bilateral Artificial Intelligence (Lampert)"}],"title":"Continual release moment estimation with differential privacy","fulldoi":"https://doi.org/10.52202/085713-1924","publisher":"Neural Information Processing Systems Foundation","month":"12","oa_version":"Published Version","conference":{"name":"NeurIPS: Neural Information Processing Systems","location":"San Diego, CA, United States"},"acknowledgement":"We thank Monika Henzinger for her valuable feedback and insightful discussions on earlier versions\r\nof this draft. We are also grateful to Mher Safaryan for his contributions to discussions on DP-Adam.\r\nAdditionally, we thank Ryan McKenna for suggesting Joint Clipping as a baseline.\r\nJalaj Upadhyay’s research was funded by the Rutgers Decanal Grant no. 302918, NSF CNS 2433628,\r\nGoogle Research Scholar Award, and Google Seed Fund Grant. A part of this work was done while\r\nvisiting the Institute of Science and Technology Austria (ISTA).\r\nNikita Kalinin’s research was funded in part by the Austrian Science Fund (FWF) [10.55776/COE12]","department":[{"_id":"GradSch"},{"_id":"ChLa"}],"alternative_title":["Advances in Neural Information Processing Systems"],"_id":"22828","quality_controlled":"1","intvolume":"        38","publication_identifier":{"isbn":["9798331338275"],"issn":["1049-5258"]},"researchdata_availability":"no","volume":38,"author":[{"first_name":"Nikita","id":"4b14526e-14d2-11ed-ba64-c14c9553d137","last_name":"Kalinin","full_name":"Kalinin, Nikita"},{"full_name":"Upadhyay, Jalaj","last_name":"Upadhyay","first_name":"Jalaj"},{"first_name":"Christoph","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","last_name":"Lampert","full_name":"Lampert, Christoph","orcid":"0000-0001-8622-7887"}],"date_updated":"2026-09-16T07:40:56Z","citation":{"chicago":"Kalinin, Nikita, Jalaj Upadhyay, and Christoph Lampert. “Continual Release Moment Estimation with Differential Privacy.” In <i>Advances in Neural Information Processing Systems</i>, 38:64002–47. Neural Information Processing Systems Foundation, 2025. <a href=\"https://doi.org/10.52202/085713-1924\">https://doi.org/10.52202/085713-1924</a>.","mla":"Kalinin, Nikita, et al. “Continual Release Moment Estimation with Differential Privacy.” <i>Advances in Neural Information Processing Systems</i>, vol. 38, Neural Information Processing Systems Foundation, 2025, pp. 64002–47, doi:<a href=\"https://doi.org/10.52202/085713-1924\">10.52202/085713-1924</a>.","ieee":"N. Kalinin, J. Upadhyay, and C. Lampert, “Continual release moment estimation with differential privacy,” in <i>Advances in Neural Information Processing Systems</i>, San Diego, CA, United States, 2025, vol. 38, pp. 64002–64047.","short":"N. Kalinin, J. Upadhyay, C. Lampert, in:, Advances in Neural Information Processing Systems, Neural Information Processing Systems Foundation, 2025, pp. 64002–64047.","ama":"Kalinin N, Upadhyay J, Lampert C. Continual release moment estimation with differential privacy. In: <i>Advances in Neural Information Processing Systems</i>. Vol 38. Neural Information Processing Systems Foundation; 2025:64002-64047. doi:<a href=\"https://doi.org/10.52202/085713-1924\">10.52202/085713-1924</a>","ista":"Kalinin N, Upadhyay J, Lampert C. 2025. Continual release moment estimation with differential privacy. Advances in Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems, Advances in Neural Information Processing Systems, vol. 38, 64002–64047.","apa":"Kalinin, N., Upadhyay, J., &#38; Lampert, C. (2025). Continual release moment estimation with differential privacy. In <i>Advances in Neural Information Processing Systems</i> (Vol. 38, pp. 64002–64047). San Diego, CA, United States: Neural Information Processing Systems Foundation. <a href=\"https://doi.org/10.52202/085713-1924\">https://doi.org/10.52202/085713-1924</a>"},"year":"2025","date_created":"2026-09-06T22:02:00Z","main_file_link":[{"open_access":"1","url":"https://doi.org/10.52202/085713-1924"}],"publication":"Advances in Neural Information Processing Systems","abstract":[{"lang":"eng","text":"We propose Joint Moment Estimation (JME), a method for continually and privately estimating both the first and second moments of a data stream with reduced noise compared to naive approaches. JME supports the matrix mechanism and exploits a joint sensitivity analysis to identify a privacy regime in which the second-moment estimation incurs no additional privacy cost, thereby improving accuracy while maintaining privacy. We demonstrate JME’s effectiveness in two applications: estimating the running mean and covariance matrix for Gaussian density estimation and model training with DP-Adam."}],"scopus_import":"1","day":"02","date_published":"2025-12-02T00:00:00Z","doi":"10.52202/085713-1924","das_tickbox":"0"},{"corr_author":"1","OA_place":"publisher","supplementarymaterial":"yes","language":[{"iso":"eng"}],"user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","page":"159016-159028","article_processing_charge":"No","oa":1,"fulldoi":"https://doi.org/10.52202/085713-4798","publication_status":"published","project":[{"_id":"d8f03aaa-b035-11f1-8588-d5147fa879e0","grant_number":"COE12","name":"Bilateral Artificial Intelligence (Lampert)"}],"title":"Setting ε is not the issue in differential privacy","type":"conference","OA_type":"free access","status":"public","publication_identifier":{"isbn":["9798331338275"],"eissn":["1049-5258"]},"researchdata_availability":"no","volume":38,"intvolume":"        38","quality_controlled":"1","alternative_title":["Advances in Neural Information Processing Systems"],"department":[{"_id":"ChLa"}],"_id":"22823","acknowledgement":"This research was funded in whole by the Austrian Science Fund (FWF) 10.55776/COE12. I thank\r\nAlberto Naibo for the discussion and his invitation to \"Logique, droit, IA : penser les algorithmes\"\r\ncolloquium. I thank Simone Bombari for the discussions. I thank EurIPS for the active poster session.","conference":{"end_date":"2025-12-07","name":"NeurIPS: Neural Information Processing Systems","start_date":"2025-12-02","location":"San Diego, CA, United States"},"publisher":"Neural Information Processing Systems Foundation","oa_version":"Published Version","month":"12","das_tickbox":"0","date_published":"2025-12-02T00:00:00Z","day":"02","doi":"10.52202/085713-4798","abstract":[{"text":"This position paper argues that setting the privacy budget in differential privacy should not be viewed as an important limitation of differential privacy compared to alternative methods for privacy-preserving machine learning. The so-called problem of interpreting the privacy budget is often presented as a major hindrance to the wider adoption of differential privacy in real-world deployments and is sometimes used to promote alternative mitigation techniques for data protection. We believe this misleads decision-makers into choosing unsafe methods. We argue that the difficulty in interpreting privacy budgets does not stem from the definition of differential privacy itself, but from the intrinsic difficulty of estimating privacy risks in context, a challenge that any rigorous method for privacy risk assessment face. Moreover, we claim that any sound method for estimating privacy risks should, given the current state of research, be expressible within the differential privacy framework or justify why it cannot.","lang":"eng"}],"scopus_import":"1","publication":"39th Conference on Neural Information Processing Systems","main_file_link":[{"open_access":"1","url":"https://doi.org/10.52202/085713-4798"}],"citation":{"ama":"Cyffers EAL. Setting ε is not the issue in differential privacy. In: <i>39th Conference on Neural Information Processing Systems</i>. Vol 38. Neural Information Processing Systems Foundation; 2025:159016-159028. doi:<a href=\"https://doi.org/10.52202/085713-4798\">10.52202/085713-4798</a>","ista":"Cyffers EAL. 2025. Setting ε is not the issue in differential privacy. 39th Conference on Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems, Advances in Neural Information Processing Systems, vol. 38, 159016–159028.","apa":"Cyffers, E. A. L. (2025). Setting ε is not the issue in differential privacy. In <i>39th Conference on Neural Information Processing Systems</i> (Vol. 38, pp. 159016–159028). San Diego, CA, United States: Neural Information Processing Systems Foundation. <a href=\"https://doi.org/10.52202/085713-4798\">https://doi.org/10.52202/085713-4798</a>","chicago":"Cyffers, Edwige Audrey Lucienne. “Setting ε Is Not the Issue in Differential Privacy.” In <i>39th Conference on Neural Information Processing Systems</i>, 38:159016–28. Neural Information Processing Systems Foundation, 2025. <a href=\"https://doi.org/10.52202/085713-4798\">https://doi.org/10.52202/085713-4798</a>.","short":"E.A.L. Cyffers, in:, 39th Conference on Neural Information Processing Systems, Neural Information Processing Systems Foundation, 2025, pp. 159016–159028.","ieee":"E. A. L. Cyffers, “Setting ε is not the issue in differential privacy,” in <i>39th Conference on Neural Information Processing Systems</i>, San Diego, CA, United States, 2025, vol. 38, pp. 159016–159028.","mla":"Cyffers, Edwige Audrey Lucienne. “Setting ε Is Not the Issue in Differential Privacy.” <i>39th Conference on Neural Information Processing Systems</i>, vol. 38, Neural Information Processing Systems Foundation, 2025, pp. 159016–28, doi:<a href=\"https://doi.org/10.52202/085713-4798\">10.52202/085713-4798</a>."},"year":"2025","date_created":"2026-09-06T22:01:59Z","date_updated":"2026-10-15T09:34:50Z","author":[{"last_name":"Cyffers","id":"20d4c299-977a-11ef-ae55-98b15ac64a57","first_name":"Edwige Audrey Lucienne","full_name":"Cyffers, Edwige Audrey Lucienne"}]}]