I am a third-year Computer Science PhD student at University of Southern California
I am very fortunate to be advised by Haipeng Luo. My research focuses on developing efficient online learning algorithms with provable guarantees.  

Research Experiences [CV]
Summer 2019: interning at Microsoft Research, Redmond (supervisor: Alekh Agarwal)
Summer 2018: interning at Yahoo Research, New York (supervisor: Alina Beygelzimer)
2015-2017: Research assistant at Academia Sinica, Taipei (supervisor: Chi-Jen Lu)   

  • Alekh Agarwal, John Langford, Chen-Yu Wei
    Federated Residual Learning  
    Preprint, 2020. 

  • Ehsan Emamjomeh-Zadeh, Chen-Yu Wei, Haipeng Luo, David Kempe
    Adversarial Online Learning with Changing Action Sets: Efficient Algorithms with Approximate Regret Bounds  
    Preprint, 2020.

  • Chen-Yu Wei, Haipeng Luo, Alekh Agarwal
    Taking a Hint: How to Leverage Loss Predictors in Contextual Bandits?  [arXiv]
    Preprint, 2020. 

  • Chen-Yu Wei, Mehdi Jafarnia-Jahromi, Haipeng Luo, Hiteshi Sharma, Rahul Jain
    Model-free Reinforcement Learning in Infinite-horizon Average-reward Markov Decision Processes  
    Preprint, 2019. 
  • James Preiss, Sébastien Arnold, Chen-Yu Wei, Marius Kloft 
    Analyzing the Variance of Policy Gradient Estimators for the Linear-Quadratic Regulator  [arXiv]
    NeurIPS Optimization Foundations for Reinforcement Learning Workshop (OptRL), 2019.
    Also appeared in Southern California Machine Learning Symposium (SoCalML), 2019.  Best poster award.  

  • Yifang Chen, Chung-Wei Lee, Haipeng Luo, Chen-Yu Wei
    A New Algorithm for Non-stationary Contextual Bandits: Efficient, Optimal, and Parameter-free  [arXiv]
    Conference on Learning Theory (COLT), 2019. 

  • Sébastien Bubeck, Yuanzhi Li, Haipeng Luo, Chen-Yu Wei
    Improved Path-length Regret Bounds for Bandits  [arXiv]
    Conference on Learning Theory (COLT), 2019. 

  • Alina Beygelzimer, Dávid Pál, Balázs Szörényi, Devanathan Thiruvenkatachari, Chen-Yu Wei, Chicheng Zhang
    Bandit Multiclass Linear Classification: Efficient Algorithms for the Separable Case  [arXiv]
    International Conference on Machine Learning (ICML), 2019. 

  • Julian Zimmert, Haipeng Luo, Chen-Yu Wei
    Beating Stochastic and Adversarial Semi-bandits Optimally and Simultaneously  [arXiv]
    International Conference on Machine Learning (ICML), 2019.  Long talk.  
  • Haipeng Luo, Chen-Yu Wei, Kai Zheng
    Efficient Online Portfolio with Logarithmic Regret  [arXiv]
    Advances in Neural Information Processing Systems (NeurIPS), 2018.  Spotlight.   

  • Haipeng Luo, Chen-Yu Wei, Alekh Agarwal, John Langford 
    Efficient Contextual Bandits in Non-stationary Worlds  [arXiv]
    Conference on Learning Theory (COLT), 2018. 

  • Chen-Yu Wei and Haipeng Luo 
    More Adaptive Algorithms for Adversarial Bandits  [arXiv]
    Conference on Learning Theory (COLT), 2018. 

  • Chen-Yu Wei, Yi-Te Hong, Chi-Jen Lu 
    Online Reinforcement Learning in Stochastic Games  [arXiv]
    Advances in Neural Information Processing Systems (NIPS), 2017.  

  • Chen-Yu Wei, Yi-Te Hong, Chi-Jen Lu 
    Tracking the Best Expert in Non-stationary Stochastic Environments  [arXiv]
    Advances in Neural Information Processing Systems (NIPS), 2016.  

Email: chenyu.wei@usc.edu