Yu-Xiang Wang's Homepage
|
|
Yu-Xiang Wang 王宇翔Eugene Aas Chair Assistant Professor Department of Computer Science, UC Santa Barbara Office: HFH-2121 E-mail: yuxiangw AT cs.ucsb.edu Yu-Xiang is pronounced approximately as ['ju:'ʃi:ʌŋ], namely, y~eu~ee - sh~ih~ah~ng . Looking for self-motivated students and postdocs. |
Welcome
Hello! Welcome to my homepage. I am a faculty member of the computer science department in UCSB. Prior joining UCSB, I was a scientist with Amazon AI in Palo Alto. Even before that I was with the Machine Learning Department in Carnegie Mellon University and had the pleasure of being jointly advised by Stephen Fienberg, Alex Smola, Ryan Tibshirani and Jing Lei.
Over the years I have worked on a diverse set of problems in the broad area of statistical machine learning, e.g., trend filtering, differential privacy, subspace clustering, large-scale learning / optimization, bandits / reinforcement learning, just to name a few. Check out my publication page for details. My most recent quests include making differential privacy practical and developing a statistical foundation for off-policy reinforcement learning.
Teaching
- (This quarter!) CS165A Artificial Intelligence (2020 Winter): [course website]
- CS292A Convex Optimization (2019 Spring): [course website]
- CS165A Artificial Intelligence (2019 Winter): [course website]
Publications, Google Scholar profile
News
- Jan 2020: Paper "Asymptotically Efficient Off-Policy Evaluation for Tabular Reinforcement Learning" accepted to AISTATS 2020. Congrats, Ming Yin!.
- Oct 2019: Visiting UC San Diego, spoke on Online Nonparametric Forecasting here.
- Sep 2019: Four papers accepted to NeurIPS 2019. Topics include Reinforcement Learning, Time Series Forecasting and Online Learning (see here) . Congrats, Dheeraj, Shiyang, Yu Bai , Tengyang!
- June 2019: Spoke in an invited session at the ICML EXPO workshop on "Machine Learning for All-Inclusive Finance" on privacy and fairness. Slides can be found here.
- June 2019: My students Dheeraj Baby and Chong Liu shared our recent work on "Online Forecasting" and Doubly Robust Crowdsourcing" at Time Series workshop and the Human In the Loop Learning (HILL) Workshop at ICML'18.
- Apr 2019: Paper "Poisson Subsampled Renyi Differential Privacy" is accepted to ICML'19. Congrats, Yuqing!
- Apr 2019: Our paper "Subsampled Renyi Differential Privacy and Analytical Moments Accountant" with Borja Balle and Shiva Kasiviswanathan received the Notable Paper Award from AISTATS'2019. Thanks anonymous reviewers and AISTATS organizers!
- Apr 2019: Spoke at the Simons Institute Workshop on Privacy and the Science of Data Analysis (see the slides here).
- Dec 2018: Spoke at UCSB Statistics Department Seminar on Off-policy evaluation and Learning. (see the slides here).
- Dec 2018: Three papers accepted to AISTATS'19. Two of them are accepted as oral presentations (top 2.5% of all submissions). (see here) .
- Dec 2018: Our ProxQuant paper is accepted to ICLR'19. Congrats, Yu Bai ! See the full paper here.
- Dec 2018: Attended NeurIPS in Montreal. Talked about "Subsampled Renyi DP" at the "Privacy Preserving Machine Learning' workshop. My intern Tengyang Xie talked about on "Marginalized Off-policy Evaluation for Reinforcement Learning" at the Causal Learning Workshop.
- Nov 2018: A few journal papers to appear: Operations Research (see here), IEEE Transaction of Information Theory (here) and the Journal of Confidentiality and Privacy (and here) .
- Nov 2018: Attended the "Privacy in Graphs" workshop in UC Santa Cruz and talked about"Per-instance Differential Privacy" (to appear in JPC).
- Nov 2018: Visited Caltech and talked about "off-policy learning" without needing to know logging probabilities.