Research map / Probabilistic and scientific modeling / Variational and generative sampling
Unifying Orthogonal Monte Carlo Methods
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 |
|---|---|---|---|
| Add multiple independent Rademacher diagonal blocks to Hadamard-Rademacher matrices (Hadamard-MultiRademacher) (ablation) | Hadamard-Rademacher random matrices · MMD on sphere, dimensionality 16 | squared MMD vs FLOPs | Helped |
| Use Kac's random walk with C=2 d log d Givens rotations instead of i.i.d. sampling for RBF kernel directions | random Fourier feature kernel approximation · Gaussian kernel pointwise approximation | empirical MSE | Helped |
| Use Hadamard-Rademacher matrices with few Givens rotations (single block) for MMD sampling | Hadamard-Rademacher · MMD on sphere, dimensionality 16 | squared MMD | Hurt |
| Use approximate OMC methods (Kac, HMD, HD, SGP, BFLY) instead of i.i.d. sampling | Gaussian kernel pointwise and matrix approximation · boston, cpu, wine, parkinson | MSE and normalized Frobenius error | Helped |
| Use Kac's random walk for RL policy gradient estimation instead of i.i.d. | Minitaur quadruped locomotion policy search · Minitaur simulator | learning of walking behaviour | Helped |
| Use Hadamard-MultiRademacher and Kac's random walk for gradient estimation in Swimmer RL | Swimmer OpenAI Gym policy search · Swimmer | learning curves (reward) | Helped |
Closest papers by meaning
- Demystifying Orthogonal Monte Carlo and Beyond (NeurIPS 2020 · similarity 0.86)
- Quadrature-based features for kernel approximation (NeurIPS 2018 · similarity 0.86)
- Sampling-based Nyström Approximation and Kernel Quadrature (ICML 2023 · similarity 0.85)
- Quasi-Monte Carlo Features for Kernel Approximation (ICML 2024 · similarity 0.85)
- Random Features for Shift-Invariant Kernels with Moment Matching (AAAI 2017 · similarity 0.85)
- Inverse M-Kernels for Linear Universal Approximators of Non-Negative Functions (NeurIPS 2024 · 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.