Research map / Optimization / Nonconvex and smooth optimization
Information Theoretic Guarantees for Empirical Risk Minimization with Applications to Model Selection and Large-Scale Optimization
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 |
|---|---|---|---|
| Parallelized averaging of 50 ERM solutions on disjoint subsets (Theorem 8) instead of single ERM on all data | ℓ2-regularized logistic regression on MiniBooNE · MiniBooNE Particle Identification | Test Error: 14.8% vs 15.3% for ERM on 1,000 examples; matches 14.8% of ERM on 50,000 | Helped |
| Parallelized averaging (K=50) runs in the time of a single 1,000-example ERM rather than 50,000-example ERM | ℓ2-regularized logistic regression on MiniBooNE · MiniBooNE Particle Identification | Time: 0.03s vs 2.93s | Helped |
| Proposed UIC criterion (adds sqrt(d/2m) penalty) instead of AIC/BIC for model selection | polynomial regression model selection · Synthetic polynomial degree selection (Figure 3) | Correct degree recovery: only UIC succeeds in estimating d* in all problems | Helped |
Closest papers by meaning
- Exactly Tight Information-theoretic Generalization Bounds via Binary Jensen-Shannon Divergence (ICML 2025 · similarity 0.89)
- Information Complexity of Stochastic Convex Optimization: Applications to Generalization, Memorization, and Tracing (ICML 2024 · similarity 0.88)
- On the Efficiency of ERM in Feature Learning (NeurIPS 2024 · similarity 0.87)
- On the Asymptotic Distribution of the Minimum Empirical Risk (ICML 2024 · similarity 0.87)
- Stability and Sharper Risk Bounds with Convergence Rate $\tilde{O}(1/n^2)$ (NeurIPS 2025 · similarity 0.87)
- The Sample Complexity of Gradient Descent in Stochastic Convex Optimization (NeurIPS 2024 · similarity 0.87)
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.