I am a fourth-year Ph.D. student in the Department of Computer Science and Technology at Tsinghua University, advised by Prof. Peng Cui and Prof. Bo Li. I got my Bachelor degree in the Department of Computer Science and Technology at Tsinghua University in 2020. I work closely with Prof. Hongseok Namkoong in the Decision, Risk, and Operations division at Columbia Business School (remote). I also collaborate with Prof. Kun Kuang at Zhejiang University.
My research interests include:
- Evaluation under Distribution Shifts.
- Measure, Explore, and Exploit Data Heterogeneity.
- Distributionally Robust Optimization.
- Applications of OOD Generalization & Heterogeneity.
I am looking for undergraduates to collaborate with. If you are interested in performance evaluation, robust learning, out-of-distribution generalization, etc. Please fill in this form.
Selected PapersThe full list of publications can be found in the publications page. * indicates equal contributions.
Evaluation under Distribution Shifts
- On the Need for a Language Describing Distribution Shifts: Illustrations on Tabular Datasets
NeurIPS 2023, Datasets and Benchmarks Track
The 37th Conference on Neural Information Processing Systems.
[paper][code] [python package] [downloads>2.3k]
- Rethinking the Evaluation Protocol of Domain Generalization
Under Review, [paper]
Explore and Exploit Data Heterogeneity
- Towards Out-of-Distribution Generalization: A Survey
Under Review at IEEE TPAMI, [paper] [website] [pageviews>19k]
- Predictive Heterogeneity: Measures and Applications
Revise & Resubmit at JMLR, [paper]
Distributionally Robust Optimization
- Distributionally Robust Learning with Stable Adversarial Training
IEEE TKDE 2022,In IEEE Transactions on Knowledge and Data Engineering.
- Distributionally Robust Optimization with Data Geometry
NeurIPS 2022,In Neural Information Processing Systems. Spotlight presentation (Top 5.2%). [paper]
- Stable Adversarial Learning under Distributional Shifts
AAAI 2021,In AAAI Conference on Artificial Intelligence. [paper] [code]
- ●Distribution robustness adversarial learning method. Issued 2022.
CN Patent: CN 112085194 B, filed August 30, 2020, and issued December 13, 2022. [certificate]
- ●Invariant learning method and device based on heterogeneity hybrid data. Issued 2023.
CN Patent: CN 113205184 B, filed April 28, 2021, and issued January 31, 2023. [certificate]
Predictive Heterogeneity: Measures and Applications
Invited by Prof. Kun Kuang's group, Lab of Artificial Intelligence, Zhejiang University, Mar. 15, 2023
Measure the Predictive Heterogeneity
Distributionally Robust Optimization with Data Geometry
Data Heterogeneity and Invariance in Out-of-Distribution Generalization
Mining the Data Heterogeneity for Out-of-Distribution Generalization
Invited by Dr Tat-Seng Chua's group, NUS-Tsinghua Extreme Search Center (NExT), June. 08, 2022 (Online)
Invited by Dr Kun Kuang's group, Lab of Artificial Intelligence, Zhejiang University, Sep. 1, 2022
Journal reviewer: Operations Research, IEEE Transactions on Multimedia
Conference reviewer / Program committee: ICLR (2024), NeurIPS (2023), ICML (2022,2023), UAI (2022,2023), AAAI (2022), IJCAI (2022,2023), CVPR (2022,2023), ICCV(2023), CoLLAs (2022,2023), AISTATS (2021,2023,2024)
Workshop reviewer / Program committee: NeurIPS DistShift (2023)
Awards and Fellowships
- Excellent Comprehensive Scholarship of Tsinghua University (for PhD. Student, 2022)
- National Scholarship for Graduate Student (2021)
- Apple Scholars in AI/ML Nomination (2021)
- Excellent Undergraduate, Tsinghua University (2020, 10%)
- TP-Link Scholarship (2019)
- Toyota Scholarship (2018)
- Excellent Academic Scholarship of Tsinghua University (2018, 2019)
- Excellent Comprehensive Scholarship of Tsinghua University (2017, 5%)
- Second-Class Freshmen Scholarship of Tsinghua University (2016~2019)
- Software Engineering (Fall 2019, 2020, 2021, 2022, Spring 2022, 2023 TA)
- Object-oriented Programming (Summer 2022, TA)
- Remove the watermark of a PDF file.
- An AI player for the Connect Four Game (ranked 100/715 in the leaderboard, Top 14%).
[Leaderboard] [Human vs AI]: find 2022AI_2020310840 and press the right button