I'm an assistant professor in Heinz College, which is part of Carnegie Mellon University, and contains the Schools of Public Policy and Information Systems.
I also have a courtesy appointment in the Department of Statistics.
My research interests are statistics and machine learning in network settings.
This includes community detection, other network models involving latent variables or unsupervised learning,
and causal inference in the presence of social network effects (i.e., interference between units).
I was a graduate student in the
Department of Electrical Engineering at Stanford University, where my advisor was
Benjamin Van Roy. I've also worked as a technical staff member at MIT Lincoln Laboratory, as a postdoc in the School of Engineering at Harvard University,
and as a visiting postdoc in the Department of Statistics at UC Berkeley.
- A Semidefinite Program for Structured Blockmodels. In submission.
- Co-clustering of non-smooth graphons.
To appear in Annals of Statistics.
- Estimation of monotone treatment effects in network experiments.
To appear in Journal of the American Statistical Association.
- Co-clustering separately exchangeable network data. (with Patrick Wolfe)
Annals of Statistics , 42(1), 29-63, 2014.
Asymptotic normality of maximum likelihood and its variational approximation for stochastic blockmodels.
(with Peter Bickel, Xiangyu Chang, and Hai Zhang) Annals of Statistics, 41(4), 1922-1943, 2013.
- Stochastic blockmodels with growing number of classes. (with Patrick Wolfe and Edoardo Airoldi)
Biometrika, 99(2), 273-284, 2012.
- Confidence sets for network structure. (with Edoardo Airoldi and Patrick Wolfe)
Statistical analysis and data mining, 4(5), 461-469, 2011.
Conference version in Neural Information Processing Systems 24, 2011. [pdf]
davidch at andrew.cmu.edu
Hamburg Hall 2101G, 5000 Forbes Avenue, Pittsburgh, PA 15213