Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees

Christian Janos Lebeda, David Erb, Tudor Cebere, Aurélien Bellet · 2026

Read Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees by Christian Janos Lebeda, David Erb, Tudor Cebere, Aurélien Bellet (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-2605-22756.txt.

Description

Lebeda and others build Lumberjack, a differentially private random forest that keeps utility where prior private forests degrade into impracticality. The method grows large random trees, then prunes them with a privacy preserving heavy hitter detector that keeps only well populated nodes. That detector's error scales with the square root of log tree height, allowing far deeper trees than earlier work. Benchmark tests show Lumberjack beats prior private random forest methods across practical privacy budgets, closing much of the gap to non private forests.

Catalog record

CREATORChristian Janos Lebeda, David Erb, Tudor Cebere, Aurélien Bellet
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90086
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.22756

Browse the full library · About the Tower of Babel