Machine Unlearning Methodology base on Stochastic Teacher Network

TL;DR

This paper introduces a stochastic teacher network approach for efficient machine unlearning, enabling rapid removal of specific data influence from deep learning models within a single epoch, matching retrained model performance.

Contribution

It proposes a novel stochastic teacher network method that significantly accelerates machine unlearning, allowing quick data influence mitigation without full retraining.

Findings

Mitigates influence of forgotten data within one epoch

Achieves model performance comparable to retraining

Effective across multiple datasets

Abstract

The rise of the phenomenon of the "right to be forgotten" has prompted research on machine unlearning, which grants data owners the right to actively withdraw data that has been used for model training, and requires the elimination of the contribution of that data to the model. A simple method to achieve this is to use the remaining data to retrain the model, but this is not acceptable for other data owners who continue to participate in training. Existing machine unlearning methods have been found to be ineffective in quickly removing knowledge from deep learning models. This paper proposes using a stochastic network as a teacher to expedite the mitigation of the influence caused by forgotten data on the model. We performed experiments on three datasets, and the findings demonstrate that our approach can efficiently mitigate the influence of target data on the model within a single epoch. This allows for one-time erasure and reconstruction of the model, and the reconstruction model achieves the same performance as the retrained model.