I'm a PhD student in Computer Science at UCLA advised by Guy Van den Broeck. Previously I studied math and computer science at GW, advised by the stupendous Poorvi Vora.
Contact me by email: obroadrick (at) ucla (dot) edu
Foundations of Probabilistic Machine Learning. What conditions make probabilistic inference tractable? Towards this question, I'm studying circuit models which serve as a unifying framework for myriad known tractable probabilistic models (e.g. Probabilistic Graphical Models). In my first paper on this topic [5], we consider a number of previously studied polynomial semantics for such circuit models and show that they are all tractable and equally expressive-efficient, i.e. that they can be reduced to each other.
Statistical Election Audits. Risk-limiting audits (RLAs) are rigorous statistical procedures used to detect errors in election results. In Professor Poorvi Vora's group, I have done work including experiments [1] and the development of a new statistical test, PROVIDENCE, the most efficient and secure ballot polling RLA known [3]. PROVIDENCE's pilot use in the US State of Rhode Island is described in their press release. PROVIDENCE is implemented in open source Arlo, the most popular election audit software in the US.
I love teaching. Since high school I have tutored for hundreds of hours at the high school and college level in mathematics, computer science, and physics. At GWU, I was a teaching assistant in Discrete Mathematics and Theory of Computation courses for several semesters.
Please give me anonymous feedback on my teaching.
"They did not die! I never said died. We lost them, I said. We lost them and we cannot find them." -Tolkien