Research map / Optimization / Nonconvex and smooth optimization
SDNA: Stochastic Dual Newton Ascent for Empirical Risk Minimization
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 the full curvature in each sampled subspace rather than only diagonal curvature | SDNA versus SDCA-like methods · ERM; convergence rate | convergence rate: 10,000 times better | Helped |
| Increase minibatch size | SDNA · mushrooms and synthetic datasets | passes over data to convergence | Helped |
| Increase minibatch size | Minibatch SDCA · mushrooms and synthetic datasets | passes over data to convergence | Hurt |
| Increase minibatch size | SDNA · mushrooms and synthetic datasets | runtime | Mixed |
| Use quadratic loss and quadratic regularizer | SDNA · ERM; expected duality gap convergence | convergence rate | Helped |
| Use SDNA rather than minibatch SDCA | ERM with uniform sampling · ERM; expected duality gap convergence | convergence rate | Helped |
Closest papers by meaning
- SDCA without Duality, Regularization, and Individual Convexity (ICML 2016 · similarity 0.90)
- Stochastic Smoothed Primal-Dual Algorithms for Nonconvex Optimization with Linear Inequality Constraints (ICML 2025 · similarity 0.86)
- Stochastic Newton Proximal Extragradient Method (NeurIPS 2024 · similarity 0.86)
- Last-Iterate Convergence of Single-Loop Stochastic Methods for Constrained Convex-Concave Minimax Problems (NeurIPS 2026 · similarity 0.86)
- Single-Loop Stochastic Algorithms for Difference of Max-Structured Weakly Convex Functions (NeurIPS 2024 · similarity 0.86)
- Two-timescale Extragradient for Finding Local Minimax Points (ICLR 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.