Research map / Optimization / Nonconvex and smooth optimization
Compact Convex Projections
Steffen Grünewälder
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 |
|---|---|---|---|
| increase CCP radius r with threshold 0.1 (ablation) | CCP regressor · Million Songs subset | least-squares error: 0.53 -> 0.17 | Improved |
| tighten CCP stopping threshold from 0.1 to 0.01 (ablation) | CCP regressor · Million Songs subset | least-squares error: 0.17 -> 0.13 | Improved |
| tighten CCP stopping threshold from 0.1 to 0.01 (ablation) | CCP regressor · Million Songs subset | run time (s): 8.89 -> 93.13 | Hurt |
| increase CCP radius r (ablation) | CCP regressor · Million Songs subset | run time (s): 1.03 -> 8.89 | Hurt |
| use CCP with r=5000 instead of GP regressor | CCP-0.1 · Million Songs subset | least-squares error: 0.121 -> 0.124 | No clear effect |
| vary GP regularization parameter (ablation) | GP regressor · Million Songs subset | least-squares error: 0.121 | No clear effect |
| use CCP regressor instead of Fast-KRR with long run time | CCP regressor · Million Songs (full) | least-squares error | Mixed |
| use CCP projection instead of cvx general solver (n=5000) | CCP projection · Gaussian process synthetic data | run time (s) | Improved |
Closest papers by meaning
- A Richer Theory of Convex Constrained Optimization with Reduced Projections and Improved Rates (ICML 2017 · similarity 0.88)
- Projection onto Minkowski Sums with Application to Constrained Learning (ICML 2019 · similarity 0.88)
- Convergences of Regularized Algorithms and Stochastic Gradient Methods with Random Projections (JMLR 2020 · similarity 0.87)
- Fast Projection Onto Convex Smooth Constraints (ICML 2021 · similarity 0.87)
- Pairwise Conditional Gradients without Swap Steps and Sparser Kernel Herding (ICML 2022 · similarity 0.87)
- A Study of First-Order Methods with a Deterministic Relative-Error Gradient Oracle (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.
