Home

Lili Su

Postdoc


MIT
Computer Science and Artificial Intelligence Laboratory (CSAIL)


e-mail: lilisu@mit.edu or lilisu3@csail.mit.edu

Office Phone: (808)-443-3198

Address:
MIT, EECS


About me

I am a Postdoc in EECS (CSAIL) at MIT, hosted by Prof. Nancy Lynch. I got my Ph.D degree from UIUC in 2017, supervised by Prof. Nitin H. Vaidya. Before that, my master work was on ordinal data processing at CSL Communication Group from Aug. 2012 to May 2014 . I received my B. M. degree in IE (Industrial Engineering) at Nankai University in 2011.


Research interests

  • Distributed computing
  • Learning
  • Fault-tolerance
  • Multi-agent optimization
  • Consensus and information aggregation

Funding

Center for Science of Information (National Science Foundation), Sept. 2016 - Present
"Defending Large-Scale Distributed Machine Learning Against Adversarial Attacks"

Selected Awards and Honors

  • Best Student Paper Award Finalist (1/3)
    The 30th International Symposium on DIStributed Computing (DISC), 2016
  • Best Student Paper Award
    The 17th International Symposium on Stabilization, Safety, and Security of Distributed Systems, 2015
  • Sundaram Seshu International Student Fellowship 
    University of Illinois at Urbana Champaign, 2016-2017
  • (Service Award) Exemplary Reviewer 
    IEEE Transactions on Communication, 2015
  • Premium Jebsen Fellowship
    Nankai University, China, 2010
  • National Fellowship
    Ministry of Education, China, 2009

Publications and Preprints

  • Y. Chen, L. Su and J. Xu,
    "Distributed Statistical Machine Learning in Adversarial Settings: Byzantine Gradient Descent,"
    submitted.
  • H. Su, L. Su, A. Dornhaus, and N. Lynch,
    "Ant-Inspired Dynamic Task Allocation via Gossiping,"
    submitted.
  • L. Su and N. H. Vaidya,
    "Asynchronous non-Bayesian learning in the presence of crash failures, "
    18th International Symposium on Stabilization, Safety, and Security of Distributed Systems (SSS 2016), Nov. 2016.
  • L. Su and N. H. Vaidya,
    "Asynchronous non-Bayesian learning in the presence of crash failures, "
    18th International Symposium on Stabilization, Safety, and Security of Distributed Systems (SSS 2016), Nov. 2016.
  • L. Su and N. H. Vaidya,
    "Robust multi-agent optimization: Coping with Byzantine agents with input redundancy,"
    18th International Symposium on Stabilization, Safety, and Security of Distributed Systems (SSS 2016), Nov. 2016.
  • L. Su and N. H. Vaidya,
    "Non-Bayesian Learning in the Presence of Byzantine Adversaries"
    International Symposium on DIStributed Computing (DISC), Sep. 2016
    Best Student Paper Award Finalist (1/3)
    Extended version available at "arXiv"
  • L. Su and N. H. Vaidya,
    (Extended) "Reaching Approximate Byzantine Consensus with Multi-hop Communication,"
    accepted for Information and Computation, Elsevier, 2016. 
  • L. Su and N. H. Vaidya,
    "Robust Push-Sum Distributed Dual Averaging: Coping with Packet-Dropping Link Failures,"
    submitted, 2016
  • L. Su and N. H. Vaidya,
    "Fault-Tolerant Multi-Agent Optimization: Optimal Distributed Algorithms,"
    ACM Symposium on Principles of Distributed Computing (PODC), July 2016
    "Full version"
  • L. Su and N. H. Vaidya,
    "Multi-Agent Optimization in the Presence of Byzantine Adversaries: Fundamental Limits,"
    IEEE American Control Conference (ACC), July 2016
  • L. Su and N. H. Vaidya,
    "Reaching Approximate Byzantine Consensus with Multi-hop Communication,"
    17th International Symposium on Stabilization, Safety, and Security of Distributed Systems (SSS 2015), Aug. 2015.
    Best Student Paper Award
    Full version appeared in "arXiv", Nov. 2014
  • L. Su and O. Milenkovic,
    "Synchronizing Rankings via Interactive Communication,"
    IEEE International Sysmposium on Information Theory (ISIT), July 2014
    Full version appeared in "arXiv", Feb. 2014
  • L. Su, F. Farnoud (Hassanzadeh), and O. Milenkovic,
    "Similarity Distances Between Permutations,"
    IEEE International Sysmposium on Information Theory (ISIT), July 2014
    Full version appeared in "arXiv", July 2013

Technical Reports

  • L. Su and N. H. Vaidya,
    "Fault-Tolerant Distributed Optimization (Part IV): Constrained Optimization with Arbitrary Directed Networks," arXiv, Nov. 2015
  • L. Su and N. H. Vaidya,
    "Fault-Tolerant Multi-Agent Optimization – Part III," arXiv, Sep. 2015
  • L. Su and N. H. Vaidya,
    "Byzantine Multi-Agent Optimization – Part II," arXiv, July 2015
  • L. Su and N. H. Vaidya,
    "Byzantine Multi-Agent Optimization – Part I," arXiv, June 2015

  • Teaching Assistant

    • ECE 526 Distributed Algorithms, Fall 2016
    • ECE 313 Probability with Engineering Applications, UIUC, Summer 2014, Fall 2014, 2012 Fall
    • GE 101 Engineering Graphics & Design, UIUC, Fall 2011, Spring 2012
    • IE 430 Engineering Economy, UIUC, Fall 2011