Research map / Reinforcement learning / MDPs and planning
Policy Newton Methods for Distortion Riskmetrics
Soumen Pachal, Mizhaan Prajit Maniyar, Prashanth L. A.
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 |
|---|---|---|---|
| use DRM (Gini deviation) objective instead of risk-neutral | ACRPN · Cliff walk | mean episodic return: -16.0 -> -13.6 | Improved |
| use DRM (Gini deviation) objective instead of risk-neutral | ACRPN · Cliff walk | minimum episodic return: -24 -> -123 | Hurt |
| use Gini deviation DRM instead of identity (risk-neutral) | DRMACRPN · Humanoid-v4 | mean episodic return: 103.0 -> 124.8 | Improved |
| use dual-power DRM instead of identity | DRMACRPN · Humanoid-v4 | mean episodic return: 103.0 -> 114.4 | Improved |
| use Gini deviation instead of dual-power DRM | DRMACRPN · Humanoid-v4 | return standard deviation: 21.1 -> 24.7 | Hurt |
| add second-order (Hessian) information to DRM policy gradient (ablation) | REINFORCE-DRM vs DRMACRPN · Cliff walk | mean episodic return: -14.1 -> -13.6 | Improved |
Closest papers by meaning
- Pessimism Meets Risk: Risk-Sensitive Offline Reinforcement Learning (ICML 2024 · similarity 0.89)
- A Reductions Approach to Risk-Sensitive Reinforcement Learning with Optimized Certainty Equivalents (ICML 2025 · similarity 0.88)
- Near-Optimal Sample Complexity for Online Constrained MDPs (NeurIPS 2025 · similarity 0.87)
- Provable Risk-Sensitive Distributional Reinforcement Learning with General Function Approximation (ICML 2024 · similarity 0.87)
- Soft Robust MDPs and Risk-Sensitive MDPs: Equivalence, Policy Gradient, and Sample Complexity (ICLR 2024 · similarity 0.87)
- Beyond CVaR: Leveraging Static Spectral Risk Measures for Enhanced Decision-Making in Distributional Reinforcement Learning (ICML 2025 · similarity 0.87)
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.
