Research map / Reinforcement learning / MDPs and planning
Learning and Planning in Average-Reward Markov Decision Processes
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 Differential Q-learning without a reference function instead of RVI Q-learning with a reference function | Average-reward control on the Access-Control Queuing Task · Access-Control Queuing Task | sensitivity of reward-rate learning to parameter choices | Helped |
| Choose a frequently occurring state–action pair as the RVI Q-learning reference instead of a rarely occurring pair | RVI Q-learning · Access-Control Queuing Task | reward-rate learning performance | Helped |
| Use the mean of action-value estimates as the RVI Q-learning reference | RVI Q-learning · Access-Control Queuing Task | reward-rate learning performance | Mixed |
| Estimate the average reward using the TD error rather than the observed-reward residual | On-policy prediction on the Two Loop task · Two Loop task | RMSVE (TVR) parameter sensitivity | Helped |
| Use Differential TD-learning for off-policy prediction | Off-policy prediction on the Two Loop task · Two Loop task | RMSVE (TVR) | Helped |
Closest papers by meaning
- Regret Analysis of Policy Gradient Algorithm for Infinite Horizon Average Reward Markov Decision Processes (AAAI 2024 · similarity 0.90)
- Learning General Parameterized Policies for Infinite Horizon Average Reward Constrained MDPs via Primal-Dual Policy Gradient Algorithm (NeurIPS 2024 · similarity 0.89)
- Planning and Learning in Average Risk-aware MDPs (NeurIPS 2025 · similarity 0.89)
- Model-Based Exploration in Monitored Markov Decision Processes (ICML 2025 · similarity 0.89)
- Optimal Single-Policy Sample Complexity and Transient Coverage for Average-Reward Offline RL (NeurIPS 2025 · similarity 0.89)
- Policy Gradient Algorithms in Average-Reward Multichain MDPs (NeurIPS 2026 · similarity 0.89)
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.