Research map / Optimization / Nonconvex and smooth optimization
Distributed Distributionally Robust Optimization with Non-Convex Objectives
Yang Jiao, Kai Yang, Dongjin Song
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 ASPIRE-EASE instead of DRFA-Prox | distributed DRO · SHL | worst-case test accuracy (%): 78.34±1.46 -> 79.16±1.13 | Improved |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO · Person Activity | worst-case test accuracy (%): 58.62±0.16 -> 59.43±0.44 | Improved |
| Use periodic communication instead of asynchronous communication (ablation) | ASPIRE-EASE · Person Activity | worst-case test accuracy (%): 59.43±0.44 -> 59.54±0.21 | No clear effect |
| Use periodic communication instead of asynchronous communication (ablation) | ASPIRE-EASE · SC-MA | worst-case test accuracy (%): 56.31±0.29 -> 56.71±0.16 | Mixed |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO · Fashion MNIST | worst-case test accuracy (%): 77.95±0.51 -> 78.82±0.07 | Improved |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO under backdoor attack · SHL | success attack rate (%): 21.23±3.63 -> 9.17±1.65 | Improved |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO under backdoor attack · Person Activity | success attack rate (%): 27.27±3.31 -> 22.36±2.33 | Improved |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO under backdoor attack · SC-MA | success attack rate (%): 30.79±3.65 -> 14.51±3.21 | Improved |
| Use ASPIRE-EASE instead of DRFA-Prox | distributed DRO under backdoor attack · Fashion MNIST | success attack rate (%): 63.24±2.47 -> 45.10±1.64 | Improved |
| Remove inactive cutting planes using EASE rather than retain them with cutting-plane method (ablation) | ASPIRE · Person Activity and SC-MA | convergence speed | Improved |
| Use asynchronous updates rather than waiting for all workers (ablation) | ASPIRE · Person Activity and SC-MA | convergence speed | Improved |
| Increase robustness budget Γ (ablation) | ASPIRE-EASE · Person Activity and SC-MA | worst-case robustness | Improved |
Closest papers by meaning
- Large-Scale Non-convex Stochastic Constrained Distributionally Robust Optimization (AAAI 2024 · similarity 0.89)
- Distributionally Robust Optimization with Bias and Variance Reduction (ICLR 2024 · similarity 0.88)
- Outlier-Robust Distributionally Robust Optimization via Unbalanced Optimal Transport (NeurIPS 2024 · similarity 0.88)
- Wasserstein Distributionally Robust Optimization through the Lens of Structural Causal Models and Individual Fairness (NeurIPS 2024 · similarity 0.87)
- Distributionally Robust Performative Optimization (NeurIPS 2025 · similarity 0.86)
- Drago: Primal-Dual Coupled Variance Reduction for Faster Distributionally Robust Optimization (NeurIPS 2024 · 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.
