Research map / Optimization / Distributed and compressed
Linear Convergent Decentralized Optimization with Compression
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 LEAD instead of compressed DGD-type baselines | Decentralized optimization · Logistic regression, heterogeneous data | convergence | Helped |
| Use LEAD instead of QDGD, DeepSqueeze, or CHOCO-SGD | AlexNet training · CIFAR10, heterogeneous data | convergence stability | Helped |
| Use LEAD instead of QDGD, DeepSqueeze, or CHOCO-SGD | AlexNet training · CIFAR10, heterogeneous data | convergence | Helped |
| Use LEAD instead of DGD | AlexNet training · CIFAR10, heterogeneous data | convergence | Mixed |
| Use LEAD instead of NIDS | Decentralized optimization · Linear regression | distance to optimum | No clear effect |
| Use LEAD instead of CHOCO-SGD or DeepSqueeze | AlexNet training · CIFAR10, homogeneous data | communication efficiency | No clear effect |
| Use infinity-norm scaling rather than other p-norm scalings in quantization (ablation) | b-bit quantization · Random vectors | compression error | Helped |
| Use infinity-norm b-bit quantization instead of top-k or random-k sparsification (ablation) | Communication compression · Random vectors at the same bit constraint | compression precision | Helped |
| Vary LEAD's alpha and gamma parameters around the selected settings (ablation) | LEAD · Linear regression | convergence | No clear effect |
| Use momentum state updates rather than directly integrating the compressed difference (ablation) | Difference-compression state update · Decentralized optimization | stability | Helped |
| Use diminishing rather than constant stepsize (ablation) | LEAD · Strongly convex smooth objectives | iteration complexity | Hurt |
Closest papers by meaning
- Scalable Distributed Stochastic Optimization via Bidirectional Compression: Beyond Pessimistic Limits (NeurIPS 2026 · similarity 0.90)
- Towards Faster Decentralized Stochastic Optimization with Communication Compression (ICLR 2025 · similarity 0.90)
- Revisiting Decentralized Online Convex Optimization with Compressed Communication (NeurIPS 2026 · similarity 0.89)
- Accelerated Methods with Compressed Communications for Distributed Optimization Problems Under Data Similarity (AAAI 2025 · similarity 0.89)
- Decentralized Optimization with Coupled Constraints (ICLR 2025 · similarity 0.88)
- EControl: Fast Distributed Optimization with Compression and Error Control (ICLR 2024 · similarity 0.88)
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.