AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Efficient nonmyopic batch active search

Shali Jiang, Gustavo Malkomes, Matthew Abbott, Benjamin Moseley, Roman Garnett

NeurIPS 2018 · Reinforcement learning › Policy gradient and actor-critic · approach: Tree search and policy optimization · 16 citations (OpenAlex)

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 changedWhereMeasuredEffect
use batch-ENS instead of greedy-batchgreedy-batch · drug discovery, b=50avg positives found: 240.1 -> 281.4Improved
use batch-ENS instead of greedy-batchgreedy-batch · drug discovery, b=100avg positives found: 208.2 -> 263.5Improved
use batch-ENS instead of uncertain-greedy batchUGB · drug discovery, b=50avg positives found: 218.8 -> 281.4Improved
use pessimistic oracle instead of optimistic oracle in sequential simulation of ENS (ablation)ss-ENS · drug discovery, b=50avg positives found: 185.3 -> 269.2Improved
use pessimistic oracle instead of sampling oracle in sequential simulation of ENS (ablation)ss-ENS · drug discovery, b=50avg positives found: 220.8 -> 269.2Improved
use ENS (nonmyopic) instead of one-step policy with pessimistic oraclesequential simulation · drug discovery, b=50avg positives found: 240.1 -> 269.2Improved
use 32 instead of 16 label samples in batch-ENS (ablation)batch-ENS · drug discovery, b=100avg positives found: 252.9 -> 263.5Improved
use 32 instead of 16 label samples in batch-ENS (ablation)batch-ENS · drug discovery, b=10avg positives found: 296.2 -> 295.5No clear effect
increase batch size from 1 to 100 for batch-ENS-32batch-ENS · drug discoveryavg positives found: 295.1 -> 263.5Hurt
use two-step instead of one-step lookahead with optimistic oraclesequential simulation · drug discovery, b=50avg positives found: 200.8 -> 172.1Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy