Research map / Learning theory and algorithms / Kernels and regression theory
Hierarchically Compositional Kernels for Scalable Nonparametric Learning
Jie Chen, Haim Avron, Vikas Sindhwani
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 PCA instead of random projection for partitioning (ablation) | hierarchical kernel · cadata | regression error curve | No clear effect |
| use PCA instead of random projection for partitioning (ablation) | hierarchical kernel · mnist | training time overhead (%): 805.67% overhead at r=58 down to 51.17% at r=937 | Hurt |
| use PCA instead of random projection for partitioning (ablation) | hierarchical kernel · mnist | partitioning time overhead (%): 3973.02% at r=58 | Hurt |
| increase rank r | PCA partitioning overhead · YearPredictionMSD | training overhead (%): 71.74% -> 2.66% | Improved |
| use hierarchical kernel instead of Nystrom, Fourier and independent kernels | approximate kernel · 7 of 8 data sets, Gaussian kernel | performance versus r | Improved |
| use hierarchical kernel instead of Nystrom, Fourier and independent kernels | approximate kernel · Gaussian kernel, training time | training time | Hurt |
| use hierarchical kernel at equal memory | approximate kernel · ijcnn1 and SUSY | performance versus memory | Improved |
| use full-rank (independent, hierarchical) instead of low-rank (Fourier, Nystrom) kernels | approximate kernel · covtype | accuracy | Improved |
| replace Gaussian base kernel with Laplace or inverse multiquadric kernel (ablation) | hierarchical kernel · all data sets | performance | No clear effect |
| use hierarchical kernel instead of Nystrom, Fourier or independent kernels | approximate kernel · cadata | stability of error curves over 30 seeds | Improved |
Closest papers by meaning
- Infinite Kernel Learning: Generalization Bounds and Algorithms (AAAI 2017 · similarity 0.88)
- Learning Data-adaptive Non-parametric Kernels (JMLR 2020 · similarity 0.88)
- Large Scale Online Kernel Learning (JMLR 2016 · similarity 0.87)
- Scalable Learning in Reproducing Kernel Krein Spaces (ICML 2019 · similarity 0.87)
- Hierarchical Kernels in Deep Kernel Learning (JMLR 2023 · similarity 0.87)
- Bridging the Gap Between Hyperdimensional Computing and Kernel Methods via the Nyström Method (AAAI 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.
