I'm a Visiting Assistant Professor in the Carnegie Mellon University Department of Statistics. I earned my PhD in 2009 in Harvard's similarly named department, and I was also a graduate associate of the Institute for Quantitative Social Science.
My current research interests include:
- Relational Data: The measurement and modelling of observations between individuals, which are generalizations of what we normally think of as networks. I work on improving complete specifications of relational ties in a system. This work makes up the bulk of my dissertation, supervised by Joseph Blitzstein.
- Models for Two-Party and Multi-Party Elections: Rather than predicting the results of elections, I've worked on methods for determining electoral fairness, in particular the notions of symmetry and responsiveness. Much of this work is with Gary King and Andrew Gelman.
- Quantitative Analysis in Sports: Sports provide a wealth of data and testable hypotheses for many current problems in stochastic modelling, as well as mechanisms for teaching these concepts to a broader audience. Plus, discovering something new about the games I love only enhances my enjoyment.
- Perfect Sampling Methods: Broadly speaking, any method by which a stochastic input can be translated to a corresponding stochastic output with no estimation or addition of noise; specifically, when the desired output is the stationary distribution of a Markov chain, the input being observations from the chain. Collaborators include Jose Blanchet and Xiao-Li Meng.
For a concise version, please see my Curriculum Vitae.
Email: act at acthomas dot ca.
|
Content copyright (c) 2009, Andrew C. Thomas.