How a convolutional neural network sees Senator Lindsey Graham's photo (R-SC) (Left). A visualization of Lindsey Graham's image after the application of  64 filters from one layer of a VGG16 architecture CNN. (Left)
How a convolutional neural network sees Senator Lindsey Graham's photo (R-SC). A visualization of one layer of a VGG16 architecture CNN.

Welcome! For the 2017-2018 academic year I am visiting Princeton University as the Microsoft Visiting Professor at the Center for Information Technology Policy. I am Assistant Professor in the departments of Public Administration and Policy and Political Science at the University of Georgia. I am also a faculty affiliate at the Institute for Artificial Intelligence

I study political methodology with a focus on text and image as data, the intersection of machine learning and causal inference, high dimensional experimental design,  Bayesian causal inference, deep learning, network analysis and algorithmic game theory.

My substantive research interests include American political institutions and bureaucracy,
public administration and policy.