Machine unlearning through model forgetting techniques: application area and emerging methods

Büşra Büyüktanır

doi:10.1007/s10586-026-06300-9

Machine unlearning through model forgetting techniques: application area and emerging methods

2026-07-24

0

PRE

AI

Gozde Karatas Baydoğmus

K

Kazım Yıldız

DOI：10.1007/s10586-026-06300-9

Original

Original request for help

Save

Abstract

En 中文

This study presents a comprehensive literature review on Machine Unlearning (MU), which aims to eliminate the influence of specific data samples from trained machine learning models under the “right to be forgotten” framework mandated by the European Union GDPR and similar regulations. This study presents a systematic literature review of MU research published between 2020 and 2025. More than 170 publications were reviewed. A taxonomy of centralized and Federated Unlearning(FU) methods is proposed, along with four main categories (Exact, Approximate, Certified, and Verifiable Unlearning). Findings are summarized across benchmark datasets, including MNIST, CIFAR-10, CIFAR-100, ImageNet, IMDB, AG News, MovieLens, as well as healthcare and finance related tabular datasets. Across these benchmarks, prior studies consistently report that approximate unlearning methods offer substantial computational advantages over full retraining while maintaining competitive model utility, whereas certified methods provide stronger formal guarantees at the cost of increased computational overhead. However, these deployments present important security risks like membership inference, data poisoning, malicious forgetting requests, data re-engineering, and model inversion attacks, as well as significant engineering challenges related to energy efficiency, communication overhead, and scalability constraints in IoT systems. The study also reviews privacy-enhancing approaches that integrate differential privacy, homomorphic encryption, secure computation, trusted computing environments, and access control mechanisms into the unlearning process. The viability of regulatory compliance is analyzed from a sectoral viewpoint in terms of ethics, justice, and accountability, especially in high-security industries like healthcare and finance. These factors highlight the necessity of scalable, verifiable, attack-resistant, and resource-efficient MU techniques for privacy-preserving AI. Proposes a unified taxonomy distinguishing centralized and Federated Unlearning, with clear separation of verification and evaluation metrics for analytical precision. Extends prior surveys by covering overlooked aspects such as the debate on approximate unlearning, certifiability, and model behavior divergence, alongside attack vectors and privacy-preserving mechanisms Provides the most comprehensive cross-domain and IoT-focused analysis to date, including smart cities, healthcare IoT, industrial IoT, and edge computing, with practical insights on adapting unlearning to resource-constrained devices, implementation in distributed IoT systems, and evaluation of computational overhead and scalability.

Keywords:

Machine Unlearning

Data Privacy

Federated Unlearning

Privacy-Preserving AI

GDPR

Security Attacks

Machine unlearning through model forgetting techniques: application area and emerging methods

Abstract

Journal

Organization

Cited Papers