Research map / Learning theory and algorithms / Kernels and regression theory
Hierarchical Kernels in Deep Kernel Learning
Wentao Huang, Houbao Lu, Haizhang Zhang
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 |
|---|---|---|---|
| add exponential-composition layers (G0 to G3) to Gaussian kernel (ablation) | Gaussian kernel SVM · CIFAR-10 | accuracy: 55.16% -> 55.55% | Improved |
| add composition layers to Gaussian kernel (ablation) | Gaussian kernel SVM · Moons | accuracy: 100.0% -> 100.0% | No clear effect |
| add composition layers to Gaussian kernel (ablation) | Gaussian kernel SVR · UCI elevators | RMSE: 0.1021 -> 0.1017 | Improved |
| add composition layers to Gaussian kernel (ablation) | Gaussian kernel SVR · UCI kin40k | RMSE: 0.0310 -> 0.0279 | Improved |
| add composition layers to Gaussian kernel (ablation) | Gaussian kernel SVR · UCI servo | RMSE: 0.1440 -> 0.1424 | Improved |
| add composition layers to exponential kernel (ablation) | l1 exponential kernel SVR · LIBSVM bodyfat | RMSE: 0.0658 -> 0.0669 | Hurt |
| add composition layers to exponential kernel (ablation) | l1 exponential kernel SVR · LIBSVM pyrim | RMSE: 0.1934 -> 0.1934 | No clear effect |
| add composition layers to exponential kernel (ablation) | l1 exponential kernel SVR · LIBSVM triazines | RMSE: 0.3365 -> 0.3365 | No clear effect |
| add layers to Gaussian kernel, large-lambda regime (ablation) | Gaussian kernel SVM · CIFAR-10 | accuracy: 29.84% -> 11.19% | Hurt |
Closest papers by meaning
- Reproducing kernel Hilbert C*-module and kernel mean embeddings (JMLR 2021 · similarity 0.87)
- A Representer Theorem for Deep Kernel Learning (JMLR 2019 · similarity 0.87)
- Hierarchically Compositional Kernels for Scalable Nonparametric Learning (JMLR 2017 · similarity 0.87)
- End-to-End Kernel Learning with Supervised Convolutional Kernel Networks (NeurIPS 2016 · similarity 0.87)
- Stochastic Kernel Regularisation Improves Generalisation in Deep Kernel Machines (NeurIPS 2024 · similarity 0.85)
- Convolutional Deep Kernel Machines (ICLR 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.
