Multi-Armed Bandit

Technique

First seen 7/10/2026

Last seen 8/19/2026

Evidence 4 chunks

NEIGHBORHOOD

No graph connections found for this entity yet. It may appear in future ingestion runs.

3 connections

The module selection in ACRM is formulated as a Multi-Armed Bandit problem.

LiFU ← implements 90% 1e

LiFU adapts seed scoring coefficients via multi-armed bandit online feedback.

The Adaptive Coverage Reward Mechanism is based on the Multi-Armed Bandit formulation for module selection.