Privacy Preservation through Practical Machine Unlearning
Robert Dilworth · 2025
Read Privacy Preservation through Practical Machine Unlearning by Robert Dilworth (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-10635.txt.
Description
Robert Dilworth, writing alone in 2025, studies machine unlearning as a way to guard privacy in trained models. Models learn from huge data sets and adapt over time. Privacy law may demand that a person's data be removed. Unlearning removes set data without a full rebuild. Dilworth tests naive retraining and an exact method built on data shards.
He weighs cost, stability, and real world fit on a spam data set. He also probes a link to learning from partly labeled data. A forest based unlearning frame stands out for keeping model quality while meeting privacy rules. Costs can be steep, though. The study backs unlearning as a key part of ethical AI.
Catalog record
| CREATOR | Robert Dilworth |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-10635