Unlearning-based sliding window for continual learning under concept drift

Michał Woźniak, Marek Klonowski, Maciej Mączynski, Bartosz Krawczyk · 2026

Read Unlearning-based sliding window for continual learning under concept drift by Michał Woźniak, Marek Klonowski, Maciej Mączynski, Bartosz Krawczyk (2026) 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-2603-14484.txt.

Description

Wozniak and co-authors use machine unlearning to track concept drift in data streams. Classic models assume a fixed data distribution, but real streams shift over time. Sliding windows adapt, yet each shift forces costly retraining from scratch. Their method removes the effect of stale samples with unlearning. Fresh data then updates the model without a full rebuild. Tests on image streams under varied drift show the method stays competitive, the authors report. They claim the first link of unlearning to drift control.

Catalog record

CREATORMichał Woźniak, Marek Klonowski, Maciej Mączynski, Bartosz Krawczyk
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37353
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.14484

Browse the full library · About the Tower of Babel