Adaptive Control Algorithms for Managing Infectious Diseases on a Network

• Activity Number: 692
• Type: Contributed
• Date/Time: Thursday, August 4, 2016 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Nicholas Meyer* and Eric Laber and Brian J. Reich and Krishna Pacifici
• Companies: North Carolina State University and North Carolina State University and North Carolina State University and North Carolina State University
• Keywords: spatio-temporal decision making ; adaptive treatment strategies ; model-based reinforcement learning

Abstract

Epidemics pose a serious and persistent global threat to human health, ecological stability, and the economy. Technological advancements have made it possible to collect, curate, and access massive amounts of data on an epidemic in real time. We develop an adaptive, model-assisted treatment allocation strategy that utilizes these data to assist policy makers by recommending locations for treatment. Initially, using observed data, the strategy estimates a low-dimensional system dynamics model to construct a low variance estimator of the optimal strategy. Using accrued data in real time, we adaptively determine the optimal switching point to use a semi-parametric estimator of the optimal treatment strategy. We demonstrate our treatment allocation strategy using a series of simulation experiments.