Debiasing Machine Unlearning with Counterfactual Examples

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Ziheng, Wang, Jia, Zhuang, Jun, Reddy, Abbavaram Gowtham, Silvestri, Fabrizio, Huang, Jin, Nag, Kaushiki, Kuang, Kun, Ning, Xin, Tolomei, Gabriele
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913327872999424
author Chen, Ziheng
Wang, Jia
Zhuang, Jun
Reddy, Abbavaram Gowtham
Silvestri, Fabrizio
Huang, Jin
Nag, Kaushiki
Kuang, Kun
Ning, Xin
Tolomei, Gabriele
author_facet Chen, Ziheng
Wang, Jia
Zhuang, Jun
Reddy, Abbavaram Gowtham
Silvestri, Fabrizio
Huang, Jin
Nag, Kaushiki
Kuang, Kun
Ning, Xin
Tolomei, Gabriele
contents The right to be forgotten (RTBF) seeks to safeguard individuals from the enduring effects of their historical actions by implementing machine-learning techniques. These techniques facilitate the deletion of previously acquired knowledge without requiring extensive model retraining. However, they often overlook a critical issue: unlearning processes bias. This bias emerges from two main sources: (1) data-level bias, characterized by uneven data removal, and (2) algorithm-level bias, which leads to the contamination of the remaining dataset, thereby degrading model accuracy. In this work, we analyze the causal factors behind the unlearning process and mitigate biases at both data and algorithmic levels. Typically, we introduce an intervention-based approach, where knowledge to forget is erased with a debiased dataset. Besides, we guide the forgetting procedure by leveraging counterfactual examples, as they maintain semantic data consistency without hurting performance on the remaining dataset. Experimental results demonstrate that our method outperforms existing machine unlearning baselines on evaluation metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2404_15760
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Debiasing Machine Unlearning with Counterfactual Examples
Chen, Ziheng
Wang, Jia
Zhuang, Jun
Reddy, Abbavaram Gowtham
Silvestri, Fabrizio
Huang, Jin
Nag, Kaushiki
Kuang, Kun
Ning, Xin
Tolomei, Gabriele
Machine Learning
Artificial Intelligence
The right to be forgotten (RTBF) seeks to safeguard individuals from the enduring effects of their historical actions by implementing machine-learning techniques. These techniques facilitate the deletion of previously acquired knowledge without requiring extensive model retraining. However, they often overlook a critical issue: unlearning processes bias. This bias emerges from two main sources: (1) data-level bias, characterized by uneven data removal, and (2) algorithm-level bias, which leads to the contamination of the remaining dataset, thereby degrading model accuracy. In this work, we analyze the causal factors behind the unlearning process and mitigate biases at both data and algorithmic levels. Typically, we introduce an intervention-based approach, where knowledge to forget is erased with a debiased dataset. Besides, we guide the forgetting procedure by leveraging counterfactual examples, as they maintain semantic data consistency without hurting performance on the remaining dataset. Experimental results demonstrate that our method outperforms existing machine unlearning baselines on evaluation metrics.
title Debiasing Machine Unlearning with Counterfactual Examples
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2404.15760