Research map / Reinforcement learning / Policy gradient and actor-critic
Asymptotic optimality of adaptive importance sampling
François Portier, Bernard Delyon
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 |
|---|---|---|---|
| weight stages using estimated importance-weight variance to downweight poor early samples | AIS · Gaussian toy integration | MSE | Improved |
| use more frequent sampling-policy updates rather than larger batches (ablation) | AIS · Gaussian toy integration | MSE | Improved |
| estimate the complete sampling variance instead of leaving it fixed (ablation) | AIS and wAIS · Gaussian toy integration, especially higher dimensions | MSE | Hurt |
| estimate only the diagonal sampling variance instead of leaving it fixed (ablation) | AIS and wAIS · Gaussian toy integration, especially higher dimensions | MSE | Hurt |
| use the target Gaussian as a fixed proposal rather than the wider proposal used by wAIS | Importance sampling · Gaussian toy integration | MSE | Mixed |
| learn the sampling policy through parametric AIS rather than knowing the target policy from the beginning | AIS · Importance sampling asymptotics | asymptotic variance | No clear effect |
Closest papers by meaning
- Adaptive Importance Sampling for Finite-Sum Optimization and Sampling with Decreasing Step-Sizes (NeurIPS 2020 · similarity 0.87)
- Adaptive Annealed Importance Sampling with Constant Rate Progress (ICML 2023 · similarity 0.87)
- Adaptive Sampling for Estimating Probability Distributions (ICML 2020 · similarity 0.85)
- Optimistic Algorithms for Adaptive Estimation of the Average Treatment Effect (ICML 2025 · similarity 0.84)
- Sample-Adaptivity Tradeoff in On-Demand Sampling (NeurIPS 2025 · similarity 0.83)
- Estimating the Permanent by Nesting Importance Sampling (ICML 2024 · 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.
