Research map / Learning theory and algorithms / Kernels and regression theory
Practical Markov Boundary Learning without Strong Assumptions
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 KMB's RKHS predictability-based learning instead of SLL | Markov boundary discovery · Synthetic mixed data with nonlinear relationships (pc=0.00, pn=1.00) | F1 score: 73.10˘5.02 -> 95.10˘0.63 | Helped |
| Use KMB's RKHS predictability-based learning instead of SLL | Markov boundary discovery · Synthetic mixed data with continuous variables and nonlinear relationships (pc=1.00, pn=1.00) | F1 score: 62.77˘5.73 -> 94.47˘0.81 | Helped |
| Replace KMB's predictability-based procedure with direct CCO minimization (ablation) | KMB · Synthetic mixed data (pc=0.00, pn=0.00) | F1 score: 96.74˘0.57 -> 93.12˘2.06 | Hurt |
| Replace KMB's predictability-based procedure with direct CCO minimization (ablation) | KMB · Synthetic mixed data (pc=0.00, pn=1.00) | F1 score: 95.10˘0.63 -> 82.73˘3.05 | Hurt |
| Increase the nonlinear-relationship proportion | KMB and comparison algorithms · Synthetic mixed data (pc=0.00) | F1 score | Mixed |
| Increase the continuous-variable proportion | KMB and comparison algorithms · Synthetic mixed data (pn=1.00) | F1 score | Mixed |
| Vary the stopping parameter k (ablation) | KMB · Synthetic mixed data | F1 score | Helped |
| Use KMB-selected features rather than all features | SVM feature selection · Real-world datasets | accuracy | Helped |
| Use KMB-selected features rather than the strongest comparison feature selector | SVM feature selection · Real-world datasets | accuracy | Helped |
Closest papers by meaning
- Contextual Feature Selection with Conditional Stochastic Gates (ICML 2024 · similarity 0.86)
- SAND: One-Shot Feature Selection with Additive Noise Distortion (ICML 2025 · similarity 0.85)
- Estimating Conditional Mutual Information for Dynamic Feature Selection (ICLR 2024 · similarity 0.84)
- Leveraging Uncertainty Estimates To Improve Classifier Performance (ICLR 2024 · similarity 0.83)
- SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning (ICML 2024 · similarity 0.83)
- Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework (NeurIPS 2025 · similarity 0.83)
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.