Research map / Learning theory and algorithms / Kernels and regression theory
Kernel Methods for Radial Transformed Compositional Data with Many Zeros
Junyoung Park, Changwon Yoon, Cheolwoo Park, Jeongyoun Ahn
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 |
|---|---|---|---|
| radial transformation (no zero substitution) instead of clr with zero replacement | kernel PCA on compositional data · real microbiome datasets (Table 2) | number of PCs needed for 50%/90% variance | Improved |
| radial transformation instead of clr with zero replacement | kernel PCA on synthetic high-dimensional compositional data · synthetic data (~69.5% zeros), Figure 6 | number of PCs to explain 50% and 90% variance | Improved |
| radial transformation instead of clr with zero replacement | kernel PCA on simulated data d=100 (qualitative) · simulated 4-cluster data, Figure 4 | visual group separation in kPCA projections | Improved |
| radial transformation instead of clr with lrDA, lrEM, or (1/2)xmin zero replacement | kernel PCA on simulated data d=15 (qualitative) · simulated data d=15, Figure 5 | visual subgroup separation | Improved |
| clr with different zero replacement methods (lrDA, lrEM, (1/2)xmin) | kernel PCA on compositional data · simulated data d=15, Figure 5 | sensitivity of projection to zero replacement | Hurt |
| radial transformation with Gaussian kernel on Gimblet/Arumugam (no clear advantage at 50% variance) | kernel PCA on Arumugam et al. 2011 data · Table 2 real data | PCs for 50% variance | No clear effect |
Closest papers by meaning
- Reproducing Kernels and New Approaches in Compositional Data Analysis (JMLR 2023 · similarity 0.93)
- Spectrally Transformed Kernel Regression (ICLR 2024 · similarity 0.83)
- On Statistical Learning Theory for Distributional Inputs (ICML 2024 · similarity 0.83)
- When does compositional structure yield compositional generalization? A kernel theory. (ICLR 2025 · similarity 0.83)
- Kernel Quantile Embeddings and Associated Probability Metrics (ICML 2025 · similarity 0.82)
- Learning in Feature Spaces via Coupled Covariances: Asymmetric Kernel SVD and Nyström method (ICML 2024 · similarity 0.81)
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.
