Hierarchical RL with a global cost controller and local marginal-value policies outperforms RMAB and heuristic baselines by 20-30% in simulated multi-cluster SARS-CoV-2 control.
Learn to intervene: An adaptive learning policy for restless bandits in application to preventive healthcare
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
The paper reformulates restless bandits as budgeted thresholding contextual bandits, proves non-asymptotic optimality for an oracle policy in a simplified setting, and shows a practical policy with sublinear regret and faster convergence than prior methods in heterogeneous environments.
Develops online WAoII and FWAoII policies for RMAB scheduling in WUR sensor networks using AoII and state estimation, reducing transmissions up to 70% versus Round Robin with acceptable RMSE.
Proposes SW-Whittle policy that achieves sub-linear dynamic regret for restless bandits with unknown non-stationary transition kernels via sliding windows and bandit-over-bandit window tuning.
citing papers explorer
-
Optimizing Resource-Constrained Non-Pharmaceutical Interventions for Multi-Cluster Outbreak Control Using Hierarchical Reinforcement Learning
Hierarchical RL with a global cost controller and local marginal-value policies outperforms RMAB and heuristic baselines by 20-30% in simulated multi-cluster SARS-CoV-2 control.
-
From Restless to Contextual: A Thresholding Bandit Reformulation For Finite-horizon Improvement
The paper reformulates restless bandits as budgeted thresholding contextual bandits, proves non-asymptotic optimality for an oracle policy in a simplified setting, and shows a practical policy with sublinear regret and faster convergence than prior methods in heterogeneous environments.
-
Fair and Efficient Scheduling for Sensor Networks via Online Whittle Index Policy
Develops online WAoII and FWAoII policies for RMAB scheduling in WUR sensor networks using AoII and state estimation, reducing transmissions up to 70% versus Round Robin with acceptable RMSE.
-
Online Learning of Whittle Indices for Restless Bandits with Non-Stationary Transition Kernels
Proposes SW-Whittle policy that achieves sub-linear dynamic regret for restless bandits with unknown non-stationary transition kernels via sliding windows and bandit-over-bandit window tuning.