Research map / Reinforcement learning / Policy gradient and actor-critic
Log-normality and Skewness of Estimated State/Action Values in Reinforcement Learning
Liangpeng Zhang, Ke Tang, Xin Yao
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 |
|---|---|---|---|
| increase observations per state-action from 20 to 200 | value iteration estimator · Chain MDP | probability of underestimation: 0.741 -> 0.603 | Improved |
| increase observations per state-action from 200 to 400 | value iteration estimator · Chain MDP | probability of underestimation: 0.603 -> 0.563 | Improved |
| increase observations per state-action from 10 to 50 | value iteration estimator · Complex maze | probability of underestimation: 0.676 -> 0.597 | Improved |
| increase observations per state-action from 100 to 200 | value iteration estimator · Complex maze | probability of underestimation: 0.563 -> 0.556 | No clear effect |
| use policy with positive-only immediate rewards (pi+) versus deterministic pi- | value iteration estimator · Chain MDP | probability of underestimation: 0.603 | Hurt |
| collect more data to push underestimation below asymptote | value iteration estimator · Complex maze | probability of underestimation: 0.5546 | No clear effect |
| test log-normal theory against empirical estimates | Theorem 3.7 log-normal distribution · Chain MDP | KS test p-value: 0.1190 | No clear effect |
Closest papers by meaning
- Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning (NeurIPS 2024 · similarity 0.86)
- Exploiting Action Impact Regularity and Exogenous State Variables for Offline Reinforcement Learning (Abstract Reprint) (AAAI 2024 · similarity 0.86)
- Enhancing Value Function Estimation through First-Order State-Action Dynamics in Offline Reinforcement Learning (ICML 2024 · similarity 0.86)
- Overestimation, Overfitting, and Plasticity in Actor-Critic: the Bitter Lesson of Reinforcement Learning (ICML 2024 · similarity 0.86)
- A Unified Principle of Pessimism for Offline Reinforcement Learning under Model Mismatch (NeurIPS 2024 · similarity 0.86)
- Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation (ICML 2025 · 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.
