Research map / Learning theory and algorithms / Prediction and decision losses
Path following algorithms for $\ell_2$-regularized $M$-estimation with approximation guarantee
Yunzhang Zhu, Renxiong Liu
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 |
|---|---|---|---|
| proposed grid selection and adaptive stopping vs equally spaced log-scale grid with fixed tolerance | Newton path following · ridge regression, n=1000, p=500 and 10000 | runtime vs suboptimality | Improved |
| proposed grid selection vs Ndiaye's adaptive unilateral scheme | Newton path following · l2-regularized logistic regression, simulated | runtime vs suboptimality | Improved |
| proposed Newton method vs baseline and Ndiaye on real data | Newton path following · a9a (LIBSVM) | runtime vs suboptimality | Improved |
| mixed method (gradient descent then Newton) vs pure gradient descent | path following · ridge regression | runtime vs suboptimality | No clear effect |
| mixed method vs pure gradient descent | path following · l2-regularized logistic regression | runtime vs suboptimality | Improved |
| Newton vs gradient descent and mixed methods as dimension grows | path following · p=500 vs p=10000 | runtime | Mixed |
| vary approximation level epsilon on iterations per grid point | path following · ridge and logistic regression | iterations per grid point | No clear effect |
| smaller initial step size alpha_1 | Algorithm 1 · theory | approximation error | Mixed |
Closest papers by meaning
- Large Stepsizes Accelerate Gradient Descent for Regularized Logistic Regression (NeurIPS 2025 · similarity 0.83)
- Kernel-based guarantees for nonlinear parametric models in Bayesian optimization (NeurIPS 2026 · similarity 0.83)
- A Statistical Theory of Regularization-Based Continual Learning (ICML 2024 · similarity 0.82)
- Sparsistency for inverse optimal transport (ICLR 2024 · similarity 0.82)
- Optimal Rates in Continual Linear Regression via Increasing Regularization (NeurIPS 2025 · similarity 0.82)
- MGDA Converges under Generalized Smoothness, Provably (ICLR 2025 · similarity 0.82)
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.
