Zhiyuan Li  (李志远)

I am a first year Ph.D student in the Department of Computer Science at Princeton University.Previously, I did my undergrad at Tsinghua University, where I was a member of Yao Class.

My research interest theoretical machine learning and convex/non-convex optimization.

  • Research Experience
    • I worked on computational economics at Tsinghua University under the supervision of Prof. Pingzhong Tang.
    • In the summer of 2016, I did a research internship at Cornell University, where I was advised by Prof. Carla Gomes and Prof. Bart Selman. We aim to develop faster algorithms with theoretical bound for computation problems in graphical models( e.g. MMAP and partition functions of MRF)
    • In the spring of 2016, I visited Cornell University. I worked on topics related to online leanring and bandits, advised by Karthik Sridharan.
    • At Tsinghua Univeristy, I worked with Prof. Wenjian Yu on problems related to tensor decompositions.

  • Papers

  • Solving Marginal MAP Problems with NP Oracles and Parity Constraints
                  Yexiang Xue, Zhiyuan Li, Stefano Ermon, Carla Gomes, Bart Selman
                  
In Proc. 30th Annual Conference on Neural Information Processing Systems(NIPS), 2016.
    Learning in Games: Robustness of Fast Convergence

            Dylan Foster, Zhiyuan Li, Thodoris Lykouris, Karthik Sridharan, Eva Tardos
            
    In Proc. 30th Annual Conference on Neural Information Processing Systems(NIPS), 2016.

    Stability of Generalized Two-sided Markets with Transaction Thresholds
            Zhiyuan Li, Yicheng Liu, Pingzhong Tang, Tingting Xu, Wei Zhan
             In Proc. 16th International Conference on Autonomous Agents and Multiagent Systems(AAMAS), 2017
             (Best Paper Award Nomination)