AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Certified Minimax Unlearning with Generalization Rates and Deletion Capacity

Jiaqi Liu, Jian Lou, Zhan Qin, Kui Ren

NeurIPS 2023 · Federated learning and privacy › Differential privacy · approach: Private online and user-level learning · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use the total Hessian, including indirect Hessian terms, rather than a conventional direct-Hessian Newton updateMinimax unlearning · SC-SC minimax unlearningunlearning certificationImproved
Avoid recomputing the updated total Hessian during unlearning (ablation)Certified Minimax Unlearning · SC-SC minimax unlearningcomputational costImproved
Use the efficient update without total-Hessian recomputation (ablation)Certified Minimax Unlearning · SC-SC minimax unlearningpopulation primal-dual riskImproved
Extend the certified method to convex-concave losses by regularizing the objective (ablation)Certified Minimax Unlearning · Convex-concave minimax unlearningdeletion capacity: c \cdot \frac{n\sqrt{\epsilon}}{(d\log(1/\delta))^{1/4}}Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy