Hi! I'm Chengtao

I work in the field of machine learning and am broadly interested in approximate inference, large-scale machine learning, Markov chains and mixing times, optimization, matrix approximations, kernel methods and probabilistic numerics.

My current research focuses on theoretical and practical aspects of probability measures induced by set functions. I am especially interested in measures that encourage diversity and non-redundancy in samples. Such probability measures arise in a surprisingly broad range of fields including physics, statistics and theoretical computer science. In machine learning, they lie at the heart of methods that crisply summarize large amounts of data, and enhance the efficiency of processing information.

Before coming to MIT, I got B.Eng. from Tsinghua University in China, advised by Jun Zhu. I have also spent great time working as an intern at research labs of Microsoft.

News

Preprints / Tech Reports

  • Column Subset Selection via Polynomial Time Dual Volume Sampling
    Chengtao Li, Stefanie Jegelka, Suvrit Sra; [Paper]
  • Batched High-dimensional Bayesian Optimization via Structural Kernel Learning
    Zi Wang, Chengtao Li, Stefanie Jegelka, Pushmeet Kohli; [Paper]
  • Fast Sampling for Strongly Rayleigh Measures with Application to Determinantal Point Processes
    Chengtao Li, Stefanie Jegelka, Suvrit Sra; [Paper]

Publications

  • Neural Program Lattices
    Chengtao Li, Daniel Tarlow, Alex Gaunt, Marc Brockschmidt, Nate Kushman; ICLR 2017, [Paper]
  • Fast Mixing Markov Chains for Strongly Rayleigh Measures, DPPs, and Constrained Sampling
    Chengtao Li, Stefanie Jegelka, Suvrit Sra; NIPS 2016, [Paper]
  • Gaussian quadrature for matrix inverse forms with applications
    Chengtao Li, Suvrit Sra, Stefanie Jegelka; ICML 2016, [Paper] [Code]
  • Fast DPP Sampling for Nystro╠łm with Application to Kernel Methods
    Chengtao Li, Stefanie Jegelka, Suvrit Sra; ICML 2016, [Paper] [Code]
  • Efficient Sampling for k-Determinantal Point Processes
    Chengtao Li, Stefanie Jegelka, Suvrit Sra; AISTATS 2016, [Paper] [Code]
  • Randomized Greedy Inference for Joint Segmentation, POS Tagging and Dependency Parsing
    Yuan Zhang, Chengtao Li, Regina Barzilay, Kareem Darwish; NAACL 2015, [Paper] [Code]
  • Bayesian Max-margin Multi-Task Learning with Data Augmentation
    Chengtao Li, Jun Zhu, Jianfei Chen; ICML 2014, [Paper]
  • Structured Output Learning with Candidate Labels for Local Parts
    Chengtao Li, Jianwen Zhang, Zheng Chen; ECML/PKDD 2013, [Paper]
  • Sentiment Topic Model with Decomposed Prior
    Chengtao Li, Jianwen Zhang, Jian-Tao Sun, Zheng Chen; SDM 2013, [Paper]

Contact Me

ctli at mit dot edu