Research map / Reinforcement learning / Visual and representation RL
Deep Reinforcement Learning in Continuous Action Spaces: a Case Study in the Game of Simulated Curling
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 |
|---|---|---|---|
| Add kernel regression continuous action search (KR-UCT style) on top of the policy-value network (ablation) | DL-UCT · Two-end curling vs DL-UCT | winning percentage: 53.23% supervised only; 66.05% after self-play | Helped |
| Add self-play reinforcement learning (5M shots) after supervised training (ablation) | KR-DL-UCT · Two-end curling vs DL-UCT | winning percentage: from 53.23% to 66.05% | Helped |
| Add long-term multi-end strategy via winning percentage table | KR-DRL · 8-end games vs AyumuGAT'17 and JiritsukunGAT'17 | winning percentage | Helped |
| Use the proposed KR-DRL-MES framework instead of hand-crafted-feature programs | curling program · 8-end games vs JiritsukunGAT'17 | winning percentage: 55.5% for JiritsukunGAT'17 as the baseline | Helped |
| Supervised-only KR-DL trained from AyumuGAT'16 shots | KR-DL · Elo rating vs GAT programs | Elo rating | No clear effect |
| Use JiritsukunGAT'16 as a strong baseline (not a change) | curling program · 8-end games | winning percentage | Mixed |
Closest papers by meaning
- The CoachAI Badminton Environment: A Novel Reinforcement Learning Environment with Realistic Opponents (Student Abstract) (AAAI 2024 · similarity 0.86)
- Geometry of Neural Reinforcement Learning in Continuous State and Action Spaces (ICLR 2025 · similarity 0.85)
- The CoachAI Badminton Environment: Bridging the Gap between a Reinforcement Learning Environment and Real-World Badminton Games (AAAI 2024 · similarity 0.85)
- In-Context Reinforcement Learning for Variable Action Spaces (ICML 2024 · similarity 0.85)
- Solving Continuous Mean Field Games: Deep Reinforcement Learning for Non-Stationary Dynamics (NeurIPS 2025 · similarity 0.84)
- Efficient Reinforcement Learning by Discovering Neural Pathways (NeurIPS 2024 · similarity 0.84)
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.