Research map / Optimization / Distributed and compressed
Gossip Dual Averaging for Decentralized Optimization of Pairwise Functions
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 a more connected network topology | Synchronous gossip dual averaging · Breast Cancer Wisconsin (AUC maximization) | convergence speed | Helped |
| Use a more connected network topology | Asynchronous gossip dual averaging · Breast Cancer Wisconsin (AUC maximization) | convergence speed | Helped |
| Use a more connected network topology | Asynchronous gossip dual averaging · Synthetic metric learning | convergence speed | Helped |
| Use asynchronous updates instead of synchronous updates | Gossip dual averaging · Breast Cancer Wisconsin (AUC maximization, Watts-Strogatz) | partial gradient computations to reach loss 0.1: 210,000 -> 25,000 | Helped |
| Use asynchronous updates instead of synchronous updates | Gossip dual averaging · Breast Cancer Wisconsin (AUC maximization) | convergence speed per iteration: roughly 5 times | Hurt |
| Use asynchronous updates instead of synchronous updates | Gossip dual averaging · Breast Cancer Wisconsin (AUC maximization) | standard deviation of node estimates | Helped |
| Replace propagated auxiliary observations with uniformly sampled observations for an unbiased partial-gradient estimate | Gossip dual averaging · Breast Cancer Wisconsin (AUC maximization, well-connected networks) | objective performance | No clear effect |
| Propagate auxiliary observations to form biased partial-gradient estimates | Gossip dual averaging · Breast Cancer Wisconsin (AUC maximization) | gradient-bias contribution | Helped |
Closest papers by meaning
- Decentralized Stochastic Optimization and Gossip Algorithms with Compressed Communication (ICML 2019 · similarity 0.89)
- Asynchronous Decentralized SGD with Quantized and Local Updates (NeurIPS 2021 · similarity 0.88)
- Unveiling the Power of Multiple Gossip Steps: A Stability-Based Generalization Analysis in Decentralized Training (NeurIPS 2025 · similarity 0.87)
- Decentralized Ranking Aggregation via Gossip: Convergence and Robustness (NeurIPS 2026 · similarity 0.87)
- Lower Bounds and Optimal Algorithms for Non-Smooth Convex Decentralized Optimization over Time-Varying Networks (NeurIPS 2024 · similarity 0.86)
- Robust Distributed Estimation: Extending Gossip Algorithms to Ranking and Trimmed Means (NeurIPS 2025 · similarity 0.86)
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.