Research map / Reinforcement learning / Visual and representation RL
Benchmarking Reinforcement Learning Algorithms for ICU Ventilator Settings: An Interpretable and Probabilistic Patient Environment for Doctor Agents
Ya-Hsi Chang, Po-Chih Kuo
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 |
|---|---|---|---|
| add action-stability penalty (ablation) | NFQ · Medium severity, MIMIC test | average reward: 8.49 -> 8.64 when removed | Hurt |
| add action-stability penalty (ablation) | NFQ · High severity, MIMIC test | time to meet extubation guidelines: 16.46 -> 16.29 hours when removed | Hurt |
| add intermediate state rewards (ablation) | CQL · High severity, MIMIC test | trajectory length: 15.46 vs. 18.33 | Hurt |
| add intermediate state rewards (ablation) | CQL · High severity, MIMIC test | time to meet: 16.19 vs. 19.92 | Hurt |
| add intermediate state rewards (ablation) | CQL · High severity, eICU OOD test | trajectory length: 17.58 vs. 22.32 | Hurt |
| add intermediate state rewards (ablation) | CQL · High severity, eICU OOD test | time to meet: 18.6 vs. 23.41 | Hurt |
| add terminal extubation reward (ablation) | CQL and BCQ · eICU OOD test | action diversity | Hurt |
| raise terminal extubation reward from 10 to 100 (ablation) | DQN · High severity, MIMIC test | trajectory length: 16.76 -> 15.39 | Improved |
| raise terminal extubation reward from 10 to 100 (ablation) | DQN · High severity, MIMIC test | time to meet: 17.15 -> 15.99 | Improved |
| raise terminal extubation reward from 10 to 100 (ablation) | agents (DQN, DDQN, BCQ, CQL) · Test and OOD test | extubation meet rate | Hurt |
| use offline RL agents instead of physician policy | DDQN · High severity, MIMIC test | extubation meet rate: 39.0 -> 95.0 | Improved |
| use offline RL agents instead of physician policy | DDQN · High severity, MIMIC test | time to meet: 55.41 -> 16.02 | Improved |
Closest papers by meaning
- Distribution-Free Uncertainty Quantification in Mechanical Ventilation Treatment: A Conformal Deep Q-Learning Framework (AAAI 2025 · similarity 0.88)
- Offline Guarded Safe Reinforcement Learning for Medical Treatment Optimization Strategies (NeurIPS 2025 · similarity 0.85)
- EpiCare: A Reinforcement Learning Benchmark for Dynamic Treatment Regimes (NeurIPS 2024 · similarity 0.85)
- Position: Reinforcement Learning in Dynamic Treatment Regimes Needs Critical Reexamination (ICML 2024 · similarity 0.84)
- Contextualized Policy Recovery: Modeling and Interpreting Medical Decisions with Adaptive Imitation Learning (ICML 2024 · similarity 0.84)
- Off-Policy Selection for Initiating Human-Centric Experimental Design (NeurIPS 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.
