Projects and Initatives
Various projects and initiatives that I've worked on (I will keep this updated). Check out the links for more details.
Princeton Data Science (PDS)
President, 2020-2021
I built and led a team of ~20 people with a mission to educate and promote data science on Princeton's campus. We organized hands-on data science workshops that reached hundreds of Princeton students, covering topics like regression, clustering, deep learning and tools like numpy, pandas, and scikit-learn. We also organized Princeton's first data science competition and hosted industry events (Uber, Amazon, Celonis, etc.). PDS is going strong under new leadership, find out more here.
A.I.D.A.N. (HackPrinceton)
Grand Prize Winner, HackPrinceton Spring 2018
We built a chatbot that can respond to user voice/typed commands in realtime and perform basic machine learning and exploratory data analysis. Requires user only to upload a CSV file of the dataset, making sophisticated data analysis accessible to users of all backgrounds.
DeepSquat (HackPrinceton)
Best Health/Fitness Hack, HackPrinceton Fall 2017
We built an app to assess squat technique, using a deep-learning based pose detection model. The app tracks keypoints as a user squats and uses these keypoints to compute metrics indicating quality of technique (e.g. back straightness).
Expressivity and Data-Dependency of Pruned Networks
Final Project, ORF 569 (Theory and Practice of Deep Learning) [Article coming soon!]
In a team of four, studied the expressivity and data-dependency of pruned networks after training and at initialization. Established limitations on the accuracy of at-init pruning methods and provided theoretical insight into these limits. Motivated path forward for more effective pruning early in training.