*reception to follow in MLH 3

Speaker: Bijaya Adhikari, Ph.D. Associate Professor of Computer Science, Emeriti-Faculty Scholar, College of Liberal Arts and Sciences, University of Iowa.

Abstract
Infectious diseases spread through networks of people and communities, raising important questions about where limited public-health resources should be placed. This talk focuses on two examples 1) choosing locations for disease surveillance and 2) deciding how to distribute vaccines. We will show how both problems can be viewed as selecting people, places, or resource levels under a limited budget. For surveillance, the benefit of each additional monitoring site often becomes smaller as more sites are added. This property, known as submodularity, allows a simple greedy strategy to perform close to the best possible solution. We will also see how a measure called `curvature' can provide even stronger guarantees for surveillance problems. Vaccine allocation is often more complicated because the effects of vaccinating different individuals can interact. In such cases, the objective may be only approximately submodular, but this weaker structure can still guide effective algorithms and provide useful performance guarantees.

Speaker's Bio
Bijaya Adhikari is an Associate Professor in the Department of Computer Science at the University of Iowa. His research focuses on AI and ML for graph-structured data, with particular emphasis on dynamical processes in large-scale networks. His recent work includes understanding and optimizing graph neural networks, developing AI-based multimorbidity indices for healthcare applications, and optimizing graph structures to limit the spread of infectious diseases. He is also a member of the Computational Epidemiology Group, where he investigates public-health challenges such as modeling disease transmission through contact networks, forecasting infectious-disease outbreaks, and designing optimal surveillance systems.