I recently got my PhD in computer science from UCSD and have the privilege of joining Instagram as a Research Scientist, Systems and Infrastructure. My doctoral advisor was Sanjoy Dasgupta. I obtained my MS in Computer Science back in 2021 from UCSD. Before that, I graduated from UC Davis in 2019 with a BS in Mathematical and Scientific Computation and a BA in Philosophy with a minor in Computer Science.

Research areas: machine learning theory, weak supervision, unsupervised/semi-supervised learning, with a focus on weighted majority vote

My research interests mostly involve ensemble methods – combining rules-of-thumb to create a classifier better than any specific part. Specifically, I’m interested in the setting where (i) the rules are weakly correlated with the ground truth (ii) there is no underlying generative process tying the rule predictions to the ground truth (iii) there is little to no labeled data. My focus is weighted majority vote under log-loss, with a focus on adversarial methods and error bounds/properties thereof.