Research & Talks

Research and Learning Interests

Mathematics, physics, computer science, electrical engineering, economics & finance, education.

Talks

Frequency Diffeomorphisms for Efficient Image Registration

Paper presentation for CS/ECE 4501/6501/6782: Machine Learning in Image Analysis at UVA · October 14, 2026 · Slides (PDF)

A presentation of other researchers’ work: the paper by Zhang, Liao, Dalca, Turk, Luo, Grant, and Golland, which represents diffeomorphisms in Fourier space to speed up image registration while keeping comparable accuracy.

Improving PILCO with a Neural Network Gaussian Process

Original research · April 30, 2023 · Slides (PDF) · Write-up

My senior year research project at Chantilly High School, on replacing the Gaussian Process in the PILCO reinforcement learning algorithm with a Neural Network Gaussian Process. I implemented PILCO from scratch, but its rollouts were nonsensical, so the proposed variant was never tested and my conclusion is a conjecture.