Probing then Editing: A Push-Pull Framework for Retain-Free Machine Unlearning in Industrial IoT

TL;DR

The paper addresses the need for selective forgetting in dynamic Industrial IoT systems without accessing retained data. It introduces Probing then Editing (PTE), a retain-free unlearning framework that first probes the forget class boundary to generate targeted edit instructions and then applies a push-pull editing regime to erase the forget class while preserving retained knowledge. The push step dismantles the forget region, while the pull step uses masked knowledge distillation from the original model to maintain retained performance, all without retain data. Experiments on CIFAR-10/100 and industrial benchmarks CWRU and SCUT-FD show PTE achieves complete forgetting with minimal loss in retained accuracy and outperforms existing retain-free methods, offering a scalable solution for real-world IIoT maintenance.

Abstract

In dynamic Industrial Internet of Things (IIoT) environments, models need the ability to selectively forget outdated or erroneous knowledge. However, existing methods typically rely on retain data to constrain model behavior, which increases computational and energy burdens and conflicts with industrial data silos and privacy compliance requirements. To address this, we propose a novel retain-free unlearning framework, referred to as Probing then Editing (PTE). PTE frames unlearning as a probe-edit process: first, it probes the decision boundary neighborhood of the model on the to-be-forgotten class via gradient ascent and generates corresponding editing instructions using the model's own predictions. Subsequently, a push-pull collaborative optimization is performed: the push branch actively dismantles the decision region of the target class using the editing instructions, while the pull branch applies masked knowledge distillation to anchor the model's knowledge on retained classes to their original states. Benefiting from this mechanism, PTE achieves efficient and balanced knowledge editing using only the to-be-forgotten data and the original model. Experimental results demonstrate that PTE achieves an excellent balance between unlearning effectiveness and model utility across multiple general and industrial benchmarks such as CWRU and SCUT-FD.

Figures (8)

• Figure 1: Machine Unlearning and Naive Retraining in Industrial IoT Scenarios
• Figure 2: Boxplot comparison.
• Figure 3: Epoch‑wise Test Forget and Retain Accuracy.
• Figure 4: Our vibration signal acquisition equipment.
• Figure 5: Original