A new research paper titled “The Forget-Set Identification Problem” has just been published in Machine Learning (Springer) by Andrea D’Angelo, Francesco Gullo, and Giovanni Stilo.
The study introduces Forget-Set Identification (ForSId) — a groundbreaking method that helps machine unlearning models determine what to forget before actually unlearning it. This innovation fills a crucial gap in the emerging field of Machine Unlearning, with implications for data privacy, ethical AI, and compliance with the “Right to be Forgotten.”
“The forget-set identification problem” arcticle available at Journal of Machine Learning - Springer