{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/differentially-private-federated-learning-a","title":"Differentially Private Federated Learning: A Client Level Perspective","arxiv_id":"1712.07557","date":"2017-12-20","proceeding":"ICLR 2019 5","authors":["Robin C. Geyer","Tassilo Klein","Moin Nabi"],"abstract":"Federated learning is a recent advance in privacy protection. In this\ncontext, a trusted curator aggregates parameters optimized in decentralized\nfashion by multiple clients. The resulting model is then distributed back to\nall clients, ultimately converging to a joint representative model without\nexplicitly having to share the data. However, the protocol is vulnerable to\ndifferential attacks, which could originate from any party contributing during\nfederated optimization. In such an attack, a client's contribution during\ntraining and information about their data set is revealed through analyzing the\ndistributed model. We tackle this problem and propose an algorithm for client\nsided differential privacy preserving federated optimization. The aim is to\nhide clients' contributions during training, balancing the trade-off between\nprivacy loss and model performance. Empirical studies suggest that given a\nsufficiently large number of participating clients, our proposed procedure can\nmaintain client-level differential privacy at only a minor cost in model\nperformance.","url_abs":"http://arxiv.org/abs/1712.07557v2","url_pdf":"http://arxiv.org/pdf/1712.07557v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/cyrusgeyer/DiffPrivate_FedLearning","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/KaiyuanZh/censor","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/SAP-samples/machine-learning-diff-private-federated-learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/SAP/machine-learning-diff-private-federated-learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/michelle0924hhx/DiffPrivate_FedLearning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"differentially-private-federated-learning-a","repo_url":"https://github.com/sapmlresearch/DiffPrivate_FedLearning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"federated-learning","task_name":"Federated Learning"},{"task_slug":"privacy-preserving","task_name":"Privacy Preserving"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1712.07557","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}