Exploring the Landscape of Machine Unlearning: A Comprehensive Survey and Taxonomy
Thanveer Shaik, Xiaohui Tao, Haoran Xie, Lin Li, Xiaofeng Zhu, Qing Li
TL;DR
This survey tackles the challenge of removing or neutralizing the influence of forgotten data in trained ML models, driven by privacy regulations and trust concerns. It presents a three-tier taxonomy—data deletion, data perturbation, and model update—along with evaluation metrics and public datasets for benchmarking. The authors discuss attack sophistication, standardization gaps, transferability, interpretability, data availability, and resource constraints, offering potential solutions and future directions, including MU in NLP, CV, and RS. The work underlines MU's potential to enable compliant, fair, and transparent AI systems while highlighting practical hurdles to real-world deployment.
Abstract
Machine unlearning (MU) is gaining increasing attention due to the need to remove or modify predictions made by machine learning (ML) models. While training models have become more efficient and accurate, the importance of unlearning previously learned information has become increasingly significant in fields such as privacy, security, and fairness. This paper presents a comprehensive survey of MU, covering current state-of-the-art techniques and approaches, including data deletion, perturbation, and model updates. In addition, commonly used metrics and datasets are also presented. The paper also highlights the challenges that need to be addressed, including attack sophistication, standardization, transferability, interpretability, training data, and resource constraints. The contributions of this paper include discussions about the potential benefits of MU and its future directions. Additionally, the paper emphasizes the need for researchers and practitioners to continue exploring and refining unlearning techniques to ensure that ML models can adapt to changing circumstances while maintaining user trust. The importance of unlearning is further highlighted in making Artificial Intelligence (AI) more trustworthy and transparent, especially with the increasing importance of AI in various domains that involve large amounts of personal user data.