Tengyuan Liang works on problems at the intersection of inference, learning, and optimization. Specifically, Liang's primary research centers around the following topics: (1) bridging the empirical and theoretical gap in modern statistical learning; (2) understanding the computational and algorithmic aspects of statistical inference; (3) exploring the role of stochasticity in solving non-convex optimization. Liang's research has appeared in journals such as The Annals of Statistics, Econometrica, the Journal of the Royal Statistical Society, the Journal of the American...