Research map / Learning theory and algorithms / Kernels and regression theory
A Linear-Time Kernel Goodness-of-Fit Test
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 |
|---|---|---|---|
| Optimize test locations and Gaussian kernel bandwidth instead of using random locations and the median heuristic (ablation) | FSSD · Gaussian vs. Laplace | test power | Helped |
| Use optimized FSSD features instead of the median-heuristic LKS test | Goodness-of-fit tests · Gaussian vs. Laplace | test power | Helped |
| Use optimized FSSD features instead of random features (ablation) | FSSD · RBM, d=50 and d_h=40 | test power | Helped |
| Use FSSD-opt instead of MMD-opt or ME-opt | Goodness-of-fit testing · RBM, d=50 and d_h=40 | test power | Helped |
| Use FSSD-opt instead of MMD-opt or ME-opt | Goodness-of-fit testing · RBM with only B_1,1 perturbed | test power | Helped |
| Use FSSD-opt instead of quadratic-time KSD | Goodness-of-fit testing · RBM with only B_1,1 perturbed | runtime | Helped |
| Use random FSSD locations instead of optimized locations (ablation) | FSSD · RBM, d=50 and d_h=10 | test power | No clear effect |
| Increase the number of test locations J (ablation) | FSSD-opt · Gaussian vs. Gaussian mixture model | test power | Mixed |
| Increase the number of test locations J (ablation) | FSSD-opt · Gaussian variance difference | test power | Helped |
| Increase the number of test locations J (ablation) | FSSD-rand · Gaussian variance difference | test power | Hurt |
| Use FSSD with random or optimized locations while the null hypothesis holds (ablation) | FSSD · Same Gaussian, p=q | false rejection rate | No clear effect |
Closest papers by meaning
- A Kernel Test of Goodness of Fit (ICML 2016 · similarity 0.92)
- KSD Aggregated Goodness-of-fit Test (NeurIPS 2022 · similarity 0.91)
- Sequential Kernel Goodness-of-fit Testing (ICML 2024 · similarity 0.88)
- Kernel Selection is Model Selection: A Unified Complexity-Penalised Approach for MMD Two-Sample Tests (NeurIPS 2026 · similarity 0.87)
- DUAL: Learning Diverse Kernels for Aggregated Two-sample and Independence Testing (NeurIPS 2025 · similarity 0.84)
- Kernel conditional tests from learning-theoretic bounds (NeurIPS 2025 · similarity 0.84)
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.