Date of Award

8-2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Industrial Engineering

Committee Member

Dr. Burak Eks¸ioglu, Committee Chair

Committee Member

Dr. Amin Khademi, Committee Co-Chair

Committee Member

Dr. Cole Smith

Committee Member

Dr. Tugce Isık

Abstract

With the motivation of the complex infectious disease control problem, we provide two different approaches to model the resource allocation problem to control an epidemic in a metapopulation. All of our models utilize a detailed stochastic simulation model that is validated with the data from the 2014 Ebola epidemic. This simulation model provides a tool for comparing the performance of different policies.

The first model defines a dynamic allocation problem, which is modeled by a Markov Decision Process, and aims to find feasible and effective quarantine policies to control an epidemic with limited resources. We assume that the populations share the capacity for quarantine, and we develop three different approximate solution methodologies to find policies to contain the disease. We provide extensive results comparing the policies obtained by different solution technique. We also develop intuitive benchmark heuristics that are used in practice. We provide insights on the prioritization and resource allocation in the setting used for this study.

The second study focuses on a resource allocation model with respect to a random budget, that is formulated by a two-stage stochastic programming model. Our optimization framework aids decision makers to open treatment units in populations and allocate beds to isolate patients from the community. Our model provides a novel methodology that uses the observations in the literature and our simulation model, and we develop functional forms to represent the cumulative number of infected individuals. Our model also is flexible to be applied in an online framework, where the decision makers can update the policies when new information becomes available. Our data-driven framework allows us to mimic the real decision making process, and we find the parameters for our optimization model based on limited data availability. We also analyze our system with respect to all available data to assess the value of perfect information. Our results show promising improvement on the number of cases when compared to the real bed allocation policies ii used in Sierra Leone during the 2014 Ebola epidemic. Similarly, this study also utilizes the detailed simulation model to give comparison of different policies.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.