Professor of Econometrics and Statistics and William Ladany Faculty Fellow
Tengyuan Liang is a Professor of Econometrics and Statistics at The University of Chicago. Professor Liang's research focuses on problems at the intersection of inference, learning, and optimization. He is the recipient of the National Science Foundation CAREER Award. His research has appeared in journals such as Econometrica, The Annals of Statistics, the Journal of the Royal Statistical Society, the Journal of the American Statistical Association, the Journal of Machine Learning Research, the SIAM Journal on Mathematics of Data Science, the Information and Inference: A Journal of the IMA, and in leading peer-reviewed machine learning venues such as the Conference on Learning Theory (COLT), the International Conference on Machine Learning (ICML), among other outlets. His research aims to: (i) bridge the empirical and theoretical gap in modern statistical learning; (ii) understand optimization and inference of over-parametrized or infinite-dimensional statistical models; (iii) explore the role of stochasticity in solving non-convex optimization.
Outside the University of Chicago, Professor Liang has experience as a Research Scientist at Yahoo! Research in New York, working on large-scale machine learning applications. He visited the Cowles Foundation for Research in Economics at Yale University as a short-term Visiting Professor in Econometrics. He served as the Associate Editor of the Operations Research, on the Editorial Board of the Journal of Machine Learning Research, and on the Senior Program Committee for the Conference on Learning Theory.
He earned a Ph.D. in Statistics from the Wharton School at the University of Pennsylvania and a B.Sc. in Mathematics from Peking University. He was awarded the J. Parker Memorial Bursk Prize and a Winkelman Fellowship from the Wharton School.
Machine learning is being tasked with an increasing number of important decisions. But the answers it generates involve a degree of uncertainty.
{PubDate}
Evaluating the performance of machine-learning tools isn’t always easily done.
{PubDate}