Latest updates: I successfully defended my Ph.D. dissertation in January, and starting in September I will be a Stein fellow in the Statistics Department at Stanford.
I am a Ph.D. candidate in the Department of Statistical Science at Duke and an alum of Amherst College. My CV is available for download here. See my publications page for a list of publications with links. My research program has two main components:
- Scalable statistical methods for high dimensional data with theoretical guarantees of robust uncertainty quantification; and
- Application of probability and stochastic process theory to develop novel statistical methods – especially Bayesian and likelihood-based methods – motivated by important problems in the sciences.
A few specific threads include
- Scaling behavior of MCMC in large samples/high dimensional settings.
- Approximations to the posterior in generalized linear models, arXiv preprint.
- Waiting times and the extremes of stochastic processes, arXiv preprint.
- Population estimation in the presence of capture heterogeneity.
- Rank identifiability of mixture models for multivariate categorical data.
I am a statistical consultant for HRDAG and have been a data science advisor/consultant for DataPad, eBay, and Treasure Data. I have also collaborated with MaxPoint Interactive, a leading quantitative advertising firm. I am also an active data blogger.
Prior to graduate school, I worked in research laboratories at Fred Hutchinson Cancer Research Center, where I studied a variety of problems in cell biology and signal transduction using Drosophila Melanogaster as a model organism. From 2005 to 2006, I was a member of the Cox lab at UCSF, where I studied Mycobacterium tuberculosis pathogenesis. From 2006 through 2010, I was a consultant at NERA Economic Consulting, where I was responsible for constructing econometric/statistical models, risk analytics, and served as a project manager on many engagements.
I spent most of 2011 living in Rio de Janeiro, Brazil and traveling in South America and Europe.