Research map / Optimization / Nonconvex and smooth optimization
A Conditional-Gradient-Based Augmented Lagrangian Framework
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 |
|---|---|---|---|
| Replace decreasing dual step-size bound with constant bound (const.) in CGAL (ablation) | CGAL · Max-cut / clustering / generalized eigenvector experiments | empirical convergence performance | Helped |
| Use CGAL instead of quadratic-penalty HCGM | CGAL vs HCGM · Max-cut G1 and G40 | convergence rate of objective residual and feasibility | Helped |
| Use CGAL instead of HCGM | k-means clustering (MNIST 1000x1000) · k-means clustering | objective residual and feasibility gap | Helped |
| Use CGAL instead of UPD for generalized eigenvector problems | CGAL vs UPD · Generalized eigenvector synthetic setups | lmo output convergence | Helped |
| Replace inexact line-search-based UPD with CGAL (no accuracy parameter tuning) | CGAL vs UPD · Max-cut G1 and G40 | convergence of objective residual and feasibility | Helped |
| Augmented Lagrangian with increasing penalty lambda_k = lambda0 sqrt(k+1) and dual step vs HCGM quadratic penalty with fixed scheme | CGAL · Generalized eigenvector problem (4 synthetic setups) | feasibility and objective residual | Helped |
Closest papers by meaning
- A Conditional Gradient Framework for Composite Convex Minimization with Applications to Semidefinite Programming (ICML 2018 · similarity 0.93)
- Conditional gradient methods for stochastically constrained convex minimization (ICML 2020 · similarity 0.90)
- An Inexact Augmented Lagrangian Framework for Nonconvex Optimization with Nonlinear Constraints (NeurIPS 2019 · similarity 0.88)
- A Study of First-Order Methods with a Deterministic Relative-Error Gradient Oracle (ICML 2024 · similarity 0.88)
- Inexact Augmented Lagrangian Methods for Conic Optimization: Quadratic Growth and Linear Convergence (NeurIPS 2024 · similarity 0.86)
- Accelerated Algorithms for Constrained Nonconvex-Nonconcave Min-Max Optimization and Comonotone Inclusion (ICML 2024 · similarity 0.85)
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.