Research map / Federated learning and privacy / Communication-efficient FL
DRIVE: One-bit Distributed Mean Estimation
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 DRIVE (Hadamard) instead of Hadamard + 1-bit stochastic quantization | 1-bit distributed mean estimation · Lognormal(0,1) DME, d=33,554,432 | NMSE: 2.9332 -> 0.0571 | Helped |
| use DRIVE (Hadamard) instead of Kashin + 1-bit SQ | 1-bit distributed mean estimation · Lognormal(0,1) DME, d=524,288 | NMSE: 0.3178 -> 0.0571 | Helped |
| replace structured Hadamard rotation with uniform random rotation (ablation) | DRIVE · Lognormal(0,1) DME, d=128 | NMSE: 0.0591 -> 0.0567 | Helped |
| use DRIVE+ (two-value K-Means reconstruction) instead of DRIVE (ablation) | DRIVE (Uniform) · Lognormal(0,1) DME, d=128 | NMSE: 0.0567 -> 0.0547 | Helped |
| use DRIVE+ (Hadamard) instead of DRIVE (Hadamard) (ablation) | DRIVE (Hadamard) · Lognormal(0,1) DME, d=128 | NMSE: 0.0591 -> 0.0591 | No clear effect |
Closest papers by meaning
- Accelerating Federated Learning with Quick Distributed Mean Estimation (ICML 2024 · similarity 0.89)
- Distributed Mean Estimation with Limited Communication (ICML 2017 · similarity 0.88)
- Universal Sample Coding (NeurIPS 2024 · similarity 0.86)
- Scalable Distributed Stochastic Optimization via Bidirectional Compression: Beyond Pessimistic Limits (NeurIPS 2026 · similarity 0.85)
- Compressed and distributed least-squares regression: convergence rates with applications to federated learning (ICML 2025 · similarity 0.85)
- Stochastic Controlled Averaging for Federated Learning with Communication Compression (ICLR 2024 · similarity 0.84)
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.