Talk, Niels Bohr Quantum Summer School, Odense, Denmark
As a part of the 2026 Niels Bohr Quantum Summer School, I took part in the student challenge. I was put into Group 7 (a.k.a The Secret Tigers), to work on a neutral atom analogue quantum computing problem. This was a lot fun with a great group of people and I learned a lot about the experimental side of things, check out my blog to learn more about my experience.
Talk, CUNY Graduate Center, Bethesda, MD, USA
This project involved applying complex-network analysis metrics to a live dataset of global conflict events to explore the dynamics of formation and counter-formation in conflict zones. I partitioned the data into friend/enemy networks based on allegiances at particular conflict events and explored scaling, betweenness, clustering, and distance metrics to better understand how conflict propagates.
Talk, CUNY Graduate Center, New York, USA
This project was the final assessment of my graduate-level cryptography course with Dr. Rosario Gennaro. This course was instrumental in shaping my future research direction. It was my first theoretical computer science course, and it could not have happened at a better time for such an exciting emerging crossover between physics and cryptography.
Talk, HIRE Columbia University Irving Medical Center, New York, USA
This project was inspired by recent graphical transformer machine-learning methods for causal discovery. The challenge for causal discovery benchmarking is the need for a known underlying causal structure to compare with. My colleague Dr. Jiheum Park found an excellent dataset on Alzheimer’s progression which has a known gold-standard causal structure. This was a useful way to test a novel AI-based methodology.
Talk, CISNET Mid-Year Conference, Seattle, USA
This talk covers a new methodology we developed to assist model decision making through visualization and information-theoretic metrics. Many microsimulation models take into account multiple time based variables and choosing which approach best fits the underlying data was largely a process of guesswork. We aimed to put this procedure on more rigorous footing to better inform other CISNET groups.
Talk, NIH, CISNET Annual Conference, Bethesda, MD, USA
This presentation encompassed the now complete, first generation gastric cancer microsimulation model. The goal of this talk was to demonstrate how well targets were being met and explain the decision-making that guided the model to its final structure.
Poster, NIH, CISNET Annual Conference, Bethesda, MD, USA
This project was presented as a part of the CISNET Junior Investigator series. The goal of this work was to use neural networks to remedy known missing data issues in a series of SEER databases. The Multilayer Perceptron model was an excellent candidate for imputation over this dataset due to the high dimensionality of underlying data. Our results were later integrated into certain CISNET models.