[arXiv]score: 0.11
Hierarchical RL Framework Models COVID Policy Under Real-World Uncertainty
June 4, 2026
A simulation of 1,000 agents uses hierarchical reinforcement learning with deep Q-networks, DDPG, and TD3 variants to optimize epidemic policy decisions while accounting for measurement noise in infection and hospitalization data and imperfect policy execution. The framework jointly models individual behavior choices and policymaker interventions.
cs.AIcs.LGcs.SI
HOW THIS AFFECTS YOU
●
researcherWorth watching for the uncertainty-aware RL formulation combining DQN with DDPG/TD3 in a multi-agent epidemic setting.
●
policyThis changes how computational policy models handle real-world execution errors and incomplete surveillance data, which are endemic to public health response.