Research map / Learning theory and algorithms / Matrix and tensor methods
Kernel-Matrix Determinant Estimates from stopped Cholesky Decomposition
Simon Bartels, Wouter Boomsma, Jes Frellsen, Damien Garreau
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 |
|---|---|---|---|
| use stopped Cholesky instead of pivoted Cholesky on hard settings | kernel log-determinant estimation · UCI datasets, RBF/OU kernels | CPU time relative to default: more than 150% -> 105% | Improved |
| increase kernel length-scale (ablation) | stopped Cholesky · PM2.5 | speed-up | Improved |
| use RBF instead of OU kernel (ablation) | stopped Cholesky · PM2.5 | speed-up | Improved |
| allow larger relative error r (above 0.1) (ablation) | stopped Cholesky · RBF with l>=exp(1) | CPU time relative to default: less than 10% | Improved |
| tiny kernel entries (below 1e-65) with row-wise implementation (ablation) | stopped Cholesky · PM2.5 and METRO, RBF l=1 or exp(1) | CPU time relative to default: above 105% | Hurt |
| use long length-scale on BANK data with OU kernel (ablation) | stopped Cholesky · BANK | fraction of data processed: more than 90% | No clear effect |
Closest papers by meaning
- Scalable Log Determinants for Gaussian Process Kernel Learning (NeurIPS 2017 · similarity 0.90)
- Variational Sparse Inverse Cholesky Approximation for Latent Gaussian Processes via Double Kullback-Leibler Minimization (ICML 2023 · similarity 0.88)
- Even Faster Kernel Matrix Linear Algebra via Density Estimation (ICML 2026 · similarity 0.88)
- Determinant Estimation under Memory Constraints and Neural Scaling Laws (ICML 2025 · similarity 0.87)
- Entrywise error bounds for low-rank approximations of kernel matrices (NeurIPS 2024 · similarity 0.85)
- Improved Algorithms for Kernel Matrix-Vector Multiplication Under Sparsity Assumptions (ICLR 2025 · 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.
