Research map / Reinforcement learning / Policy gradient and actor-critic
Efficient nonmyopic batch active search
Shali Jiang, Gustavo Malkomes, Matthew Abbott, Benjamin Moseley, Roman Garnett
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 batch-ENS instead of greedy-batch | greedy-batch · drug discovery, b=50 | avg positives found: 240.1 -> 281.4 | Improved |
| use batch-ENS instead of greedy-batch | greedy-batch · drug discovery, b=100 | avg positives found: 208.2 -> 263.5 | Improved |
| use batch-ENS instead of uncertain-greedy batch | UGB · drug discovery, b=50 | avg positives found: 218.8 -> 281.4 | Improved |
| use pessimistic oracle instead of optimistic oracle in sequential simulation of ENS (ablation) | ss-ENS · drug discovery, b=50 | avg positives found: 185.3 -> 269.2 | Improved |
| use pessimistic oracle instead of sampling oracle in sequential simulation of ENS (ablation) | ss-ENS · drug discovery, b=50 | avg positives found: 220.8 -> 269.2 | Improved |
| use ENS (nonmyopic) instead of one-step policy with pessimistic oracle | sequential simulation · drug discovery, b=50 | avg positives found: 240.1 -> 269.2 | Improved |
| use 32 instead of 16 label samples in batch-ENS (ablation) | batch-ENS · drug discovery, b=100 | avg positives found: 252.9 -> 263.5 | Improved |
| use 32 instead of 16 label samples in batch-ENS (ablation) | batch-ENS · drug discovery, b=10 | avg positives found: 296.2 -> 295.5 | No clear effect |
| increase batch size from 1 to 100 for batch-ENS-32 | batch-ENS · drug discovery | avg positives found: 295.1 -> 263.5 | Hurt |
| use two-step instead of one-step lookahead with optimistic oracle | sequential simulation · drug discovery, b=50 | avg positives found: 200.8 -> 172.1 | Hurt |
Closest papers by meaning
- Training Greedy Policy for Proposal Batch Selection in Expensive Multi-Objective Combinatorial Optimization (ICML 2024 · similarity 0.87)
- Geometric Active Exploration in Markov Decision Processes: the Benefit of Abstraction (ICML 2024 · similarity 0.86)
- Efficient Biological Data Acquisition through Inference Set Design (ICLR 2025 · similarity 0.86)
- ALINE: Joint Amortization for Bayesian Inference and Active Data Acquisition (NeurIPS 2025 · similarity 0.85)
- MOOSE-Chem2: Exploring LLM Limits in Fine-Grained Scientific Hypothesis Discovery via Hierarchical Search (NeurIPS 2025 · similarity 0.85)
- SeeA*: Efficient Exploration-Enhanced A* Search by Selective Sampling (NeurIPS 2024 · similarity 0.85)
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.
