Research map / Learning theory and algorithms / Matrix and tensor methods
Principal Component Hierarchy for Sparse Quadratic Programs
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 principal-component dual program rather than warm start | Sparse regression · Synthetic regression, ρ=0.9 | MSE: 3.714 -> 1.895 | Helped |
| Use the principal-component dual program rather than BR | Sparse regression · Synthetic regression, ρ=0.9 | MSE: 3.666 -> 1.895 | Helped |
| Use the principal-component dual program rather than warm start | Sparse regression · Synthetic regression, ρ=0.8 | MSE: 1.988 -> 1.863 | Helped |
| Use the principal-component dual program rather than screening | Sparse regression · SuperConductivity (SC), in-sample | MSE: 1.256\times 10^{-2} -> 1.262\times 10^{-2} | Hurt |
| Use the principal-component dual program rather than screening | Sparse regression · Crime (CR), in-sample | MSE: 2.760\times 10^{-2} -> 2.775\times 10^{-2} | Hurt |
| Use DP screening rather than screening | Sparse regression · Facebook (FB) | reduced problem size: 49 -> 11 | Helped |
| Use DP screening rather than screening | Sparse regression · UJIndoor (UJ) | reduced problem size: 465 -> 16 | Helped |
| Reduce η from 10 to 0.1 | All evaluated sparse-regression methods · Synthetic regression | MSE: at least 5 times larger | Hurt |
| Increase the principal-component subspace dimension (ablation) | DP · SuperConductivity (SC) | MSE | Helped |
| Vary the principal-component subspace dimension (ablation) | BR · SuperConductivity (SC) | MSE: k_{BR}^\star =13 | Mixed |
Closest papers by meaning
- Exploiting Numerical Sparsity for Efficient Learning : Faster Eigenvector Computation and Regression (NeurIPS 2018 · similarity 0.89)
- Local Linear Convergence of Gradient Methods for Subspace Optimization via Strict Complementarity (NeurIPS 2022 · similarity 0.88)
- Upper bounds for Model-Free Row-Sparse Principal Component Analysis (ICML 2020 · similarity 0.87)
- Scalable First-order Method for Certifying Optimal k-Sparse GLMs (ICML 2025 · similarity 0.87)
- Provable Acceleration of Nesterov's Accelerated Gradient for Asymmetric Matrix Factorization and Linear Neural Networks (NeurIPS 2024 · similarity 0.86)
- Efficient Sparse PCA via Block-Diagonalization (ICLR 2025 · 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.