Learning-Augmented Moment Estimation on Time-Decay Models

Soham Nagawanshi, Shalini Panthangi, Chen Wang, David P. Woodruff, Samson Zhou · 2026

Read Learning-Augmented Moment Estimation on Time-Decay Models by Soham Nagawanshi, Shalini Panthangi, Chen Wang, David P. Woodruff, Samson Zhou (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-2603-02488.txt.

Description

Soham Nagawanshi, Shalini Panthangi, Chen Wang, and colleagues add machine learning help to streaming algorithms in this 2026 paper. Streaming algorithms summarize huge data flows in tiny memory. Learning-augmented versions use a model's predictions to beat classic space limits. Unequal weights break that progress, though, as with sliding windows where old data must expire.

Their oracle flags the heavy hitters, the most frequent items in a data set. With that help, they build algorithms for several core problems in time-decay settings. Targets include norm and moment estimation, frequency estimation, and rectangular variants. Theory is paired with tests on real and synthetic data. Results show the predicted space savings hold up in practice, even as old items fade.

Catalog record

CREATORSoham Nagawanshi, Shalini Panthangi, Chen Wang, David P. Woodruff, Samson Zhou
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87712
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.02488

Browse the full library · About the Tower of Babel