Research map / Reinforcement learning / MDPs and planning
Reinforcement Learning When All Actions Are Not Always Available
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 SAS policy gradient or SAS natural policy gradient instead of SAS-Q-learning | San Francisco routing · San Francisco routing | performance | No clear effect |
| Use SAS policy gradient or SAS natural policy gradient instead of SAS-Q-learning | Maze robot locomotion · Maze | performance | Mixed |
| Use SAS policy gradient or SAS natural policy gradient instead of SAS-Q-learning | Product recommender system · Product recommender system | performance | Helped |
| Use SAS-Q-learning instead of the proposed policy-gradient methods | San Francisco routing · San Francisco routing with varying action availability | performance | Helped |
| Use state and available-action-set baseline combination with adaptive weights | SAS-PG · San Francisco routing, maze, and product recommender system | gradient-estimate variance | Helped |
Closest papers by meaning
- Stochastic Variance-Reduced Policy Gradient (ICML 2018 · similarity 0.89)
- Real-Time Reinforcement Learning (NeurIPS 2019 · similarity 0.88)
- Learning Optimal Deterministic Policies with Stochastic Policy Gradients (ICML 2024 · similarity 0.87)
- Beyond Stationarity: Convergence Analysis of Stochastic Softmax Policy Gradient Methods (ICLR 2024 · similarity 0.86)
- Solving Robust Markov Decision Processes: Generic, Reliable, Efficient (AAAI 2025 · similarity 0.86)
- Last-Iterate Global Convergence of Policy Gradients for Constrained Reinforcement Learning (NeurIPS 2024 · similarity 0.86)
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.