Research map / Learning theory and algorithms / Prediction and decision losses
Stein Points
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 Monte Carlo optimisation of the Stein Point objective instead of Nelder-Mead or grid search | Stein Points (greedy/herding) · Gaussian mixture test | Wasserstein distance W_P | Helped |
| Under a low computational budget, Monte Carlo-optimised Stein Points versus greedy Stein Points at larger budgets | Stein Points · Gaussian mixture test | Wasserstein distance W_P vs log neval | Mixed |
| Use Stein Points (greedy) instead of MED, SVGD for posterior approximation at fixed cost | Stein Points · GP regression and IGARCH | Wasserstein distance W_P | Mixed |
| Use Monte Carlo or greedy Stein Points depending on budget vs MED | MED baseline · GP regression and IGARCH | numerical stability | Hurt |
| Replace the herding objective with the greedy objective that adds a k0(x,x)/2 regularizer | Stein Greedy · Theory (discrepancy bound) | discrepancy rate | Helped |
| Use inverse log kernel k2 or IMQ score kernel k3 instead of IMQ | KSD methods · Gaussian mixture test | Wasserstein distance W_P | Mixed |
Closest papers by meaning
- Stein Point Markov Chain Monte Carlo (ICML 2019 · similarity 0.91)
- Bayesian Posterior Approximation via Greedy Particle Optimization (AAAI 2019 · similarity 0.88)
- Gradient-Free Kernel Stein Discrepancy (NeurIPS 2023 · similarity 0.87)
- Stein Variational Gradient Descent as Moment Matching (NeurIPS 2018 · similarity 0.87)
- Random Feature Stein Discrepancies (NeurIPS 2018 · similarity 0.87)
- The Polynomial Stein Discrepancy for Assessing Moment Convergence (ICML 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.