Ever since I joined North Carolina State University, the primary focus for the application of my research has been in the field of healthcare. Along with a team of researchers, we employ a mix of data analysis, simulation and optimization-based tools to help inform public policy, efficiently allocate resources, and develop and validate intervention techniques to improve long-term and overall health outcomes. Some of the projects that I am currently assisting with include
- Working with a detailed simulation-based model of individuals’ choice and response to public health type interventions based on clinical, observational data to create a validated model of colorectal cancer progression.
- Integrating electronic health records (EHR) and clinical expertise to provide a framework to diagnose and accurately classify patients within the sepsis spectrum, and develop and validate intervention policies that inform sepsis treatment decisions and resource allocation.
- Developing a queuing-based model that informs mental health officials regarding the necessary interventions that need to be taken in order to reduce patient wait times, thus having a direct impact of long-term mental health of the patient. You can read more about this here.
Apart from these, I have, in the past, been actively associated with projects that help in
- Identifying the best locations in which to place sensors in a smart-home environment, in order to effectively monitor the health of geriatric patients.
- Studying the network of an online smoking cessation group with the intention of identifying network structures that assist smokers looking to quit.
Sequential Assignment Under Uncertainty
The stochastic sequential assignment problem (SSAP), studies the allocation of available distinct workers with deterministic values to sequentially-arriving tasks with stochastic parameters so as to maximize the expected total reward obtained from the assignments. The goal is to find a balance between the frameworks of exploration vs exploitation. During the course of my time at the University at Buffalo, I worked on a few variations of this problem, including
- The case wherein the quality of the workers to be assigned is not known. We assumed knowledge regarding pairwise comparisons between the workers and used this info to inform assignment to the sequentially arriving tasks.
- The case wherein the workers being assigned to the tasks may retire at any stage. This work was carried out at the University of Florida in association with the Air Force Research Lab.
- Currently, we are attempting to use the framework of Sequential Stochastic Assignment in the realm of rank-aggregation algorithms. We posit the use of SSAP in validating the output obtained from using rank-aggregation algorithms.
Social Network Analysis
Social Network Analysis research studies the behavior of individuals embedded in social networks, and the how structure of such networks affect socio-economic events and processes. The field of SNA was the first step in my academic career in Operations Research. In Feb 2014, I attended a conference in St. Petersburg, FL of International Network for Social Network Analysis which gave me an amazing insight into the world of Social Network Analysis. The work I have done in this field is primarily introductory and include the following.
- Attempted to bridge the gap between Exponential Random Graph modeling (ERGM) and Longitudinal Network Formation modeling. The project also attempted at overcoming some fundamental challenges of ERGMs – including the issue of degeneracy.
- Assisted with a project involved in the studying of user-engagement patterns in an Online Health Social Network (smoking cessation forum) in order to develop a framework for prescriptive applied research.