{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T22:06:29Z","timestamp":1782857189476,"version":"3.54.5"},"reference-count":27,"publisher":"Association for Computing Machinery (ACM)","issue":"7","license":[{"start":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T00:00:00Z","timestamp":1782777600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["Commun. ACM"],"published-print":{"date-parts":[[2026,7,1]]},"abstract":"<jats:p>In the context of machine unlearning, the primary challenge lies in effectively removing traces of private data from trained models while maintaining model performance and security against privacy attacks like membership inference attacks. Traditional gradient-based unlearning methods often rely on extensive historical gradients, which becomes impractical with high unlearning ratios and may reduce the effectiveness of computation. To address these limitations, we introduce Mini-Unlearning, based on our theoretical analysis. Our analysis suggests that unlearned parameters correlate with retrained parameters through contraction mapping, a finding that guides our algorithm design. Our method, Mini-Unlearning, utilizes a minimal subset of historical gradients and leverages this contraction mapping to facilitate scalable, efficient unlearning. This lightweight, scalable method significantly enhances model accuracy and strengthens resistance to membership inference attacks. Our experiments demonstrate that Mini-Unlearning works well under higher unlearning ratios and outperforms existing techniques in accuracy and security, offering a promising solution for applications requiring robust unlearning capabilities.<\/jats:p>","DOI":"10.1145\/3793867","type":"journal-article","created":{"date-parts":[[2026,6,25]],"date-time":"2026-06-25T20:06:51Z","timestamp":1782418011000},"page":"74-81","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Machine Unlearning with Minimal Gradient Dependence for High Unlearning Ratios"],"prefix":"10.1145","volume":"69","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9067-2270","authenticated-orcid":false,"given":"Tao","family":"Huang","sequence":"first","affiliation":[{"name":"Minjiang University, School of Computer and Data Science, Fuzhou, Fujian, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-9526-9706","authenticated-orcid":false,"given":"Ziyang","family":"Chen","sequence":"additional","affiliation":[{"name":"Minjiang University, School of Computer and Data Science, Fuzhou, Fujian, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-4305-6175","authenticated-orcid":false,"given":"Jiayang","family":"Meng","sequence":"additional","affiliation":[{"name":"Renmin University of China, Beijing, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2735-2359","authenticated-orcid":false,"given":"Xu","family":"Yang","sequence":"additional","affiliation":[{"name":"Minjiang University, School of Computer and Data Science, Fuzhou, Fujian, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1996-2635","authenticated-orcid":false,"given":"Guolong","family":"Zheng","sequence":"additional","affiliation":[{"name":"Minjiang University, School of Computer and Data Science, Fuzhou, Fujian, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7351-5724","authenticated-orcid":false,"given":"Xun","family":"Yi","sequence":"additional","affiliation":[{"name":"RMIT University, Computer Science, Melbourne, Victoria, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5512-114X","authenticated-orcid":false,"given":"Ibrahim","family":"Khalil","sequence":"additional","affiliation":[{"name":"RMIT University, Melbourne, Victoria, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,6,30]]},"reference":[{"key":"e_1_2_2_2_2","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms5308"},{"key":"e_1_2_2_3_2","doi-asserted-by":"publisher","DOI":"10.1016\/S0168-1699(99)00046-0"},{"key":"e_1_2_2_4_2","doi-asserted-by":"publisher","DOI":"10.5555\/3119411.3119541"},{"key":"e_1_2_2_5_2","doi-asserted-by":"crossref","unstructured":"Chen M. et al. When machine unlearning jeopardizes privacy. In Proceedings of the 2021 ACM SIGSAC Conf. on Computer and Communications Security\u00a0(2021) 896\u2013911.","DOI":"10.1145\/3460120.3484756"},{"key":"e_1_2_2_6_2","doi-asserted-by":"crossref","unstructured":"Doon R. Rawat T. K. and Gautam S. Cifar-10 classification using deep convolutional neural network. In 2018 IEEE Punecon. IEEE\u00a0(2018) 1\u20135.","DOI":"10.1109\/PUNECON.2018.8745428"},{"key":"e_1_2_2_7_2","unstructured":"Ergen T. and Pilanci M. Revealing the structure of deep neural networks via convex duality. In Intern. Conf. on Machine Learning. PMLR\u00a0(2021) 3004\u20133014."},{"key":"e_1_2_2_8_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i11.29092"},{"key":"e_1_2_2_9_2","doi-asserted-by":"crossref","unstructured":"Graves L. Nagisetty V. and Ganesh V. Amnesiac machine learning. In Proceedings of the AAAI Conf. on Artificial Intelligence\u00a035\u00a0(2021) 11516\u201311524.","DOI":"10.1609\/aaai.v35i13.17371"},{"key":"e_1_2_2_10_2","unstructured":"Guo C. Goldstein T. Hannun A. and Maaten L. V. D. Certified data removal from machine learning models. arXiv preprint (2019) \u00a0arXiv:1911.03030."},{"key":"e_1_2_2_11_2","doi-asserted-by":"crossref","unstructured":"He K. Zhang X. Ren S. and Sun J. Deep residual learning for image recognition. In Proceedings of the IEEE Conf. on Computer Vision and Pattern Recognition. IEEE\u00a0(2016) 770\u2013778.","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_2_2_12_2","unstructured":"Izzo Z. Smart M. A. Chaudhuri K. and Zou J. Approximate data deletion from machine learning models. In Intern. Conf. on Artificial Intelligence and Statistics. PMLR\u00a0(2021) 2008\u20132016."},{"key":"e_1_2_2_13_2","doi-asserted-by":"crossref","unstructured":"Kim J. and Woo S. S. Efficient two-stage model retraining for machine unlearning. In Proceedings of the IEEE\/CVF Conf. on Computer Vision and Pattern Recognition\u00a0(2022) 4361\u20134369.","DOI":"10.1109\/CVPRW56347.2022.00482"},{"key":"e_1_2_2_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"e_1_2_2_15_2","unstructured":"Liu Y. et al. Learn to forget: Machine unlearning via neuron masking. arXiv preprint (2020) arXiv:2003.10933."},{"key":"e_1_2_2_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3679014"},{"key":"e_1_2_2_17_2","doi-asserted-by":"crossref","unstructured":"Mehta R. Pal S. Singh V. and Ravi S. N. Deep unlearning via randomized conditionally independent hessians. In Proceedings of the IEEE\/CVF conf. on Computer Vision and Pattern Recognition\u00a0(2022) 10422\u201310431.","DOI":"10.1109\/CVPR52688.2022.01017"},{"key":"e_1_2_2_18_2","unstructured":"Neel S. Roth A. and Sharifi-Malvajerdi S. Descent-to-delete: Gradient-based methods for machine unlearning. In Algorithmic Learning Theory. PMLR\u00a0(2021) 931\u2013962."},{"key":"e_1_2_2_19_2","first-page":"4","volume-title":"NIPS workshop on deep learning and unsupervised feature learning","author":"Netzer Y.","year":"2011","unstructured":"Netzer, Y. et al. Reading digits in natural images with unsupervised feature learning. In NIPS workshop on deep learning and unsupervised feature learning, Vol. 2011. Granada 2011), 4."},{"key":"e_1_2_2_20_2","unstructured":"Nguyen T. T. et al. A survey of machine unlearning. arXiv preprint (2022) \u00a0arXiv:2209.02299."},{"key":"e_1_2_2_21_2","first-page":"18075","article-title":"Remember what you want to forget: Algorithms for machine unlearning","volume":"34","author":"Sekhari A.","year":"2021","unstructured":"Sekhari, A., Acharya, J., Kamath, G., and Suresh, A. T. Remember what you want to forget: Algorithms for machine unlearning. Advances in Neural Information Processing Systems 34 (2021), 18075\u201318086.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_2_2_22_2","doi-asserted-by":"crossref","unstructured":"Shokri R. Stronati M. Song C. and Shmatikov V. Membership inference attacks against machine learning models. In 2017 IEEE Symp. on Security and Privacy. IEEE\u00a0(2017) 3\u201318.","DOI":"10.1109\/SP.2017.41"},{"key":"e_1_2_2_23_2","unstructured":"Ullah E. and Arora R. From adaptive query release to machine unlearning. In Intern. Conf. on Machine Learning. PMLR\u00a0(2023) 34642\u201334667."},{"key":"e_1_2_2_24_2","unstructured":"Warnecke A. Pirch L. Wressnegger C. and Rieck K. Machine unlearning of features and labels. arXiv preprint (2021) \u00a0arXiv:2108.11577."},{"key":"e_1_2_2_25_2","unstructured":"Wu Y. Dobriban E. and Davidson S. DeltaGrad: Rapid retraining of machine learning models. In Intern. Conf. on Machine Learning. PMLR\u00a0(2020) 10355\u201310366."},{"key":"e_1_2_2_26_2","doi-asserted-by":"publisher","DOI":"10.1007\/s42979-023-01767-4"},{"key":"e_1_2_2_27_2","unstructured":"Zhao B. Mopuri K. R. and Bilen H. iDLG: Improved deep leakage from gradients. arXiv preprint (2020) arXiv:2001.02610."},{"key":"e_1_2_2_28_2","article-title":"Deep leakage from gradients","volume":"32","author":"Zhu L.","year":"2019","unstructured":"Zhu, L., Liu, Z., and Han, S. Deep leakage from gradients. Advances in Neural Information Processing Systems 32 (2019).","journal-title":"Advances in Neural Information Processing Systems"}],"container-title":["Communications of the ACM"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3793867","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T21:14:15Z","timestamp":1782854055000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3793867"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,30]]},"references-count":27,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2026,7,1]]}},"alternative-id":["10.1145\/3793867"],"URL":"https:\/\/doi.org\/10.1145\/3793867","relation":{},"ISSN":["0001-0782","1557-7317"],"issn-type":[{"value":"0001-0782","type":"print"},{"value":"1557-7317","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,30]]},"assertion":[{"value":"2024-07-25","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-06-30","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}