Heterogeneous Decentralized Machine Unlearning with Seed Model Distillation
It addresses the need for scalable unlearning functionality in personalized IoT services, which is incremental over existing centralized methods.
The paper tackles the problem of efficiently removing user data from trained models in decentralized IoT settings, achieving state-of-the-art performance as demonstrated by experiments on three real-world datasets.
As some recent information security legislation endowed users with unconditional rights to be forgotten by any trained machine learning model, personalized IoT service providers have to put unlearning functionality into their consideration. The most straightforward method to unlearn users' contribution is to retrain the model from the initial state, which is not realistic in high throughput applications with frequent unlearning requests. Though some machine unlearning frameworks have been proposed to speed up the retraining process, they fail to match decentralized learning scenarios. In this paper, we design a decentralized unlearning framework called HDUS, which uses distilled seed models to construct erasable ensembles for all clients. Moreover, the framework is compatible with heterogeneous on-device models, representing stronger scalability in real-world applications. Extensive experiments on three real-world datasets show that our HDUS achieves state-of-the-art performance.