2025-01-14-yamashita25a.md

title	One-Shot Machine Unlearning with Mnemonic Code
booktitle	Proceedings of the 16th Asian Conference on Machine Learning
year	2025
volume	260
series	Proceedings of Machine Learning Research
month	0
publisher	PMLR
pdf	https://raw.githubusercontent.com/mlresearch/v260/main/assets/yamashita25a/yamashita25a.pdf
url	https://proceedings.mlr.press/v260/yamashita25a.html
software	https://github.com/tomyamkum/OneShotMU-with-MNCode
openreview	JQ7Ri3ccx6
abstract	Ethical and privacy issues inherent in artificial intelligence (AI) applications have been a growing concern with the rapid spread of deep learning. Machine unlearning (MU) is the research area that addresses these issues by making a trained AI model forget about undesirable training data. Unfortunately, most existing MU methods incur significant time and computational costs for forgetting. Therefore, it is often difficult to apply these methods to practical datasets and sophisticated architectures, e.g., ImageNet and Transformer. To tackle this problem, we propose a lightweight and effective MU method. Our method identifies the model parameters sensitive to the forgetting targets and adds perturbation to such model parameters. We identify the sensitive parameters by calculating the Fisher Information Matrix (FIM). This approach does not require time-consuming additional training for forgetting. In addition, we introduce class-specific random signals called mnemonic code to reduce the cost of FIM calculation, which generally requires the entire training data and incurs significant computational costs. In our method, we train the model with mnemonic code; when forgetting, we use a small number of mnemonic codes to calculate the FIM and get the effective perturbation for forgetting. Comprehensive experiments demonstrate that our method is faster and better at forgetting than existing MU methods. Furthermore, we show that our method can scale to more practical datasets and sophisticated architectures.
layout	inproceedings
issn	2640-3498
id	yamashita25a
tex_title	One-Shot Machine Unlearning with Mnemonic Code
firstpage	255
lastpage	270
page	255-270
order	255
cycles	false
bibtex_editor	Nguyen, Vu and Lin, Hsuan-Tien
editor	given family Vu Nguyen given family Hsuan-Tien Lin
bibtex_author	Yamashita, Tomoya and Yamada, Masanori and Shibata, Takashi
author	given family Tomoya Yamashita given family Masanori Yamada given family Takashi Shibata
date	2025-01-14
address
container-title	Proceedings of the 16th Asian Conference on Machine Learning
genre	inproceedings
issued	date-parts 2025 1 14
extras	label link Supplementary PDF https://raw.githubusercontent.com/mlresearch/v260/main/assets/assets/yamashita25a/yamashita25a-supp.pdf