基于忘却序列的机器学习遗忘
机器学习
2024-10-22 v2 计算机视觉与模式识别
摘要
机器遗忘（MU）正成为实现“被遗忘权利”的有前景范例，其中可消除任何选定数据点的训练痕迹，同时在遗忘后保持对一般测试样本的模型实用性。随着遗忘研究的发展，许多根本性问题仍未得到解答：不同样本在被遗忘方面是否存在不同难度？遗忘顺序由其各自难度决定是否影响遗忘算法的性能？本文识别影响遗忘难度和遗忘算法性能的关键因素。我们发现，具有更高隐私风险的样本更可能被遗忘，这表明遗忘难度在不同样本之间存在差异，从而激发了更精确的遗忘模式。基于这一洞察，我们提出了通用的遗忘框架，称为 RSU，其包含 Ranking 模块和 SeqUnlearn 模块。
引用
@article{arxiv.2410.06446,
title = {Machine Unlearning in Forgettability Sequence},
author = {Junjie Chen and Qian Chen and Jian Lou and Xiaoyu Zhang and Kai Wu and Zilong Wang},
journal= {arXiv preprint arXiv:2410.06446},
year = {2024}
}
备注
The senior authors of the draft are not fully convinced that the novelty is significant enough for this submission compared to the latest research progress in this area. Additionally, the senior authors have identified writing issues. Based on these two reasons, we have decided to withdraw the draft from arXiv