Research map / Reinforcement learning / Visual and representation RL
Dynamic Measurement Scheduling for Event Forecasting using Deep RL
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 an LSTM RNN instead of RF | MIMIC3 24-hour mortality prediction · MIMIC3 test set | AUC: 0.935 -> 0.950 | Helped |
| Use an LSTM RNN instead of LR | MIMIC3 24-hour mortality prediction · MIMIC3 test set | AUPR: 0.752 -> 0.803 | Helped |
| Schedule measurements sequentially with a state-dependent DQN rather than following physician scheduling | MIMIC3 ICU mortality forecasting · MIMIC3 | action cost at the same information gain: 31% | Helped |
| Schedule measurements sequentially with a state-dependent DQN rather than following physician scheduling | MIMIC3 ICU mortality forecasting · MIMIC3 | information gain at the same action cost: 3 times | Helped |
| Model measurement rewards independently instead of using sequential joint scheduling (ablation) | DQN measurement scheduling · MIMIC3 | information gain: 30% | Hurt |
| Model measurement rewards independently instead of using sequential joint scheduling (ablation) | DQN measurement scheduling · MIMIC3 | action cost reduction: 12% | Hurt |
| Measure every available feature | MIMIC3 mortality forecasting policy · MIMIC3 | information gain | Hurt |
| Dynamically condition measurement choices on patient state rather than use fixed heuristic schedules | Dueling DQN measurement scheduling · Simulation | predictive probability of detecting disease at the same action frequency | Helped |
Closest papers by meaning
- Position: Reinforcement Learning in Dynamic Treatment Regimes Needs Critical Reexamination (ICML 2024 · similarity 0.83)
- EpiCare: A Reinforcement Learning Benchmark for Dynamic Treatment Regimes (NeurIPS 2024 · similarity 0.82)
- Not Only Where, But When: Temporal Scheduling for RLVR (NeurIPS 2026 · similarity 0.82)
- Value-Based Deep RL Scales Predictably (ICML 2025 · similarity 0.82)
- Offline Guarded Safe Reinforcement Learning for Medical Treatment Optimization Strategies (NeurIPS 2025 · similarity 0.81)
- Adam on Local Time: Addressing Nonstationarity in RL with Relative Adam Timesteps (NeurIPS 2024 · similarity 0.81)
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.