John D. Lafferty

 


John Lafferty
Louis Block Professor
Department of Statistics
Department of Computer Science
Physical Sciences Division
University of Chicago


I am a professor at the University of Chicago, with a joint appointment in Statistics (primary) and Computer Science, and a faculty member of the College of the University of Chicago, currently serving on the College Council. I am also an Adjoint Professor at the Toyota Technological Institute at Chicago. My research is in statistical machine learning, with a current focus on computational and statistical aspects of nonparametric methods, high-dimensional data, graphical models, and document and text analysis. I'm part of the Computational and Applied Mathematics Initiative (CAMI) at the University of Chicago. A short NSF-style biographical sketch is here.

Recent activities


Rina Foygel Barber, Tracy Ke, and I organize a research and reading group at the University of Chicago called HELIOS (High-dimensional Estimation, Learning, Inference and Optimization in Statistics).

Dave Blei and I have launched the Hopper project to study statistical machine learning algorithms for extracting representations of text and mathematical expressions and building new search tools for the published scientific literature. Hopper is supported by a grant from the Alfred P. Sloan Foundation.

During summer 2016 I'll be teaching a minicourse at Columbia on high dimensional statistical learning.

A report was prepared from a 2014 NRC workshop we organized on training students to extract value from big data.

The University of Chicago Department of Statistics is celebrating its 60th anniversary this spring. We organized a workshop in connection with the 2016 Bahadur lectures.

In September 2016 we are running a workshop at the Becker-Friedman Institute for Research in Economics on the theme Machine Learning: What's in it for Economics?

Recent publications and preprints


  • Local minimax complexity of stochastic convex optimization
    Yuancheng Zhu, Sabyasachi Chatterjee, John Duchi and John Lafferty
    arxiv:1605.07596

  • Convergence analysis for rectangular matrix completion using Burer-Monteiro factorization and gradient descent
    Qinqing Zheng and John Lafferty
    arxiv:1605.07051

  • Denoising flows on trees
    Sabyasachi Chatterjee and John Lafferty
    arxiv:1602.08048

  • Adaptive risk bounds in unimodal regression
    Sabyasachi Chatterjee and John Lafferty
    arxiv:1512.02956

  • Quantized nonparametric estimation
    Yuancheng Zhu and John Lafferty
    arxiv:1503.07368

  • Faithful variable selection for high dimensional convex regression
    Min Xu, Minhua Chen and John Lafferty
    arxiv:1411.1805

Teaching


Research support


The research of my group is currently supported by grants from the National Science Foundation, the Alfred P. Sloan Foundation, the Office of Naval Research, DARPA, Amazon AWS in Education Machine Learning, and Facebook (AI/ML).