arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.17993cs.LGstat.ML
收藏

具有长期性能保证的序列遗忘的反事实认证

Counterfactual Certification for Sequential Unlearning with Long-Run Performance Guarantees

DI ENS, École normale supérieure, Université PSL, CNRS(法国高等师范学院,巴黎文理研究大学,法国国家科学研究中心);CMAP, École polytechnique, Institut Polytechnique de Paris(巴黎综合理工学院,巴黎理工学院)

Hanna Benarroch, Jamal Atif, Olivier Cappé

AI总结:

该研究针对序列删除请求的机器遗忘问题,提出固定微调预算下的反事实认证方法,可使性能损失一致有界,在FEMNIST数据集上表现优于现有序列方法及训练时差分隐私。

摘要:

认证式机器遗忘确保被遗忘模型与未使用待遗忘数据重新训练的模型之间的接近性。将该框架扩展到序列删除请求(通常为不同大小的组,如单个用户或协同用户的所有记录)颇具挑战性。对于基于保留集微调的遗忘方法,随着遗忘请求数量增加,会导致性能下降或计算量增大。针对独立的连续请求,我们针对被删除组从未参与的反事实遗忘轨迹对每次模型发布进行认证。在每次删除的固定微调预算下,我们证明对于平滑强凸目标,该方法可使性能损失在请求数量上保持一致有界。在FEMNIST数据集上使用多类逻辑回归的实验表明,所得算法比现有序列方法需要更少的校正步骤,且优于训练时差分隐私,尤其在组大小异质时表现更优。

英文摘要:

Certified machine unlearning ensures the proximity between the unlearned model and a model retrained without the data to be forgotten. Extending this framework to sequential deletion requests, typically arriving as groups of variable size (e.g., all records of one user or of coordinated users), is challenging. For unlearning methods based on fine-tuning on the retain sets, this leads to either degrading performance or increasing computation as the number of unlearning requests grows. For independent successive requests, we instead certify each model release against a counterfactual unlearning trajectory in which the group being deleted never took part. With a fixed fine-tuning budget per deletion, we show that, for smooth strongly convex objectives, this approach keeps the performance loss bounded, uniformly in the number of requests. Experiments with multiclass logistic regression on FEMNIST show that the resulting algorithm needs fewer corrective steps than existing sequential methods and outperforms training-time differential privacy, particularly when group sizes are heterogeneous.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑