Research map / Federated learning and privacy / Differential privacy
Practical Locally Private Heavy Hitters
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use TreeHist instead of Bassily and Smith | Local heavy hitters · Worst-case theoretical performance | server time: ˜O(n5/2) -> ˜O(n) | Helped |
| Use TreeHist instead of Bassily and Smith | Local heavy hitters · Worst-case theoretical performance | user time: ˜O(n3/2) -> ˜O(1) | Helped |
| Use Bitstogram instead of Bassily and Smith | Local heavy hitters · Worst-case theoretical performance | worst-case error: O (√(n log(d))) -> O (√(n log(d))) | No clear effect |
| Use TreeHist instead of Bassily and Smith | Local heavy hitters · Worst-case theoretical performance | worst-case error: O (√(n log(d))) -> O (√(n log(n) log(d))) | Hurt |
| Apply TreeHist to NLTK Brown corpus using threshold=15√n | TreeHist · NLTK Brown corpus | recall (TPR): 0.86 (σ = 0.05) | Helped |
Closest papers by meaning
- Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees (NeurIPS 2026 · similarity 0.84)
- Sort, Partition, Randomize: Optimal Binary Hypothesis Testing under Local Differential Privacy (NeurIPS 2026 · similarity 0.83)
- Faster Differentially Private Top-$k$ Selection: A Joint Exponential Mechanism with Pruning (NeurIPS 2024 · similarity 0.83)
- Locally Optimal Private Sampling: Beyond the Global Minimax (NeurIPS 2025 · similarity 0.82)
- Lightweight Protocols for Distributed Private Quantile Estimation (ICML 2025 · similarity 0.81)
- Individualized Privacy Accounting via Subsampling with Applications in Combinatorial Optimization (ICML 2024 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.