Research map / Reinforcement learning / Offline RL
Weighted model estimation for offline model-based reinforcement learning
Toru Hishinuma, Kei Senda
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 |
|---|---|---|---|
| weighted model estimation (Algorithm 1) instead of unweighted ERM | pendulum model · Pendulum policy evaluation | estimated discounted return (true -15.7): -80.7 -> -17.8 | Improved |
| add weight α = 0.2 (ablation) | Algorithm 3 · Walker2d-medium-expert | normalized score: 59.5 ± 49.4 -> 108.2 ± 0.5 | Improved |
| add weight α = 0.2 (ablation) | Algorithm 3 · Walker2d-medium | normalized score: 81.7 ± 1.2 -> 60.7 ± 29.0 | Hurt |
| add weight α = 0.2 (ablation) | Algorithm 3 · HalfCheetah-medium-expert | normalized score: 73.9 ± 24.2 -> 85.7 ± 21.6 | Improved |
| add weight α = 0.2 (ablation) | Algorithm 3 · Hopper-medium-replay | normalized score: 97.2 ± 10.9 -> 104.0 ± 3.2 | Improved |
| add weight α = 0.2 (ablation) | Algorithm 3 · Hopper-medium-expert | normalized score: 109.3 ± 1.1 -> 104.9 ± 10.1 | No clear effect |
| add weight α = 0.2 (ablation) | Algorithm 3 · HalfCheetah-medium-replay | normalized score: 72.1 ± 1.4 -> 65.5 ± 6.4 | Hurt |
| add weight α = 0.2 (ablation) | Algorithm 3 · HalfCheetah-medium | normalized score: 75.7 ± 1.5 -> 73.1 ± 5.2 | No clear effect |
| add weight α = 0.2 (ablation) | Algorithm 3 · Walker2d-medium-replay | normalized score: 80.7 ± 3.1 -> 82.7 ± 3.3 | No clear effect |
| add weight α = 0.2 (ablation) | Algorithm 3 · Hopper-random | normalized score: 30.2 ± 4.4 -> 32.7 ± 0.5 | Improved |
| simplified weighted loss (Algorithm 2) versus ERM start of iteration | Algorithm 2 · Pendulum | weighted loss | Improved |
| weighted model estimation (α = 0 MOPO-like) versus original MOPO | Algorithm 3 with α = 0 · Hopper-medium | normalized score: 28.0 ± 12.4 -> 100.9 ± 2.7 | Improved |
Closest papers by meaning
- Model Selection for Off-policy Evaluation: New Algorithms and Experimental Protocol (NeurIPS 2025 · similarity 0.87)
- SUMO: Search-Based Uncertainty Estimation for Model-Based Offline Reinforcement Learning (AAAI 2025 · similarity 0.87)
- Constrained Latent Action Policies for Model-Based Offline Reinforcement Learning (NeurIPS 2024 · similarity 0.87)
- Dynamic Uncertainty Estimation for Offline Reinforcement Learning (AAAI 2025 · similarity 0.87)
- Model-based Offline Reinforcement Learning with Lower Expectile Q-Learning (ICLR 2025 · similarity 0.86)
- OCEAN-MBRL: Offline Conservative Exploration for Model-Based Offline Reinforcement Learning (AAAI 2024 · 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.
