Liyan Xie
 |
Liyan Xie |
Contact
E-Mail: xieliyan@cuhk.edu.cn
Address: Office 420D, Daoyuan Building, CUHK-Shenzhen
I am actively looking for Postdoc and Ph.D. students. Please send me an email with your CV if you are interested.
Research Interest
My research interest lies in the interface of statistics, optimization, and machine learning, inspired by important applications.
Statistical Inference: Time series analysis, sequential hypothesis testing and change detection.
Optimization and Machine Learning: Diffusion models, synthetic data generation, data-driven decision-making under uncertainty, distributionally robust methods, and adversarial robustness.
Applications: Healthcare (wearable sensors and electronic health record data), spatio-temporal data analysis, community detection, and manufacturing quality control.
Recent News
January 2024, I will be organizing a session on ‘‘change point detection and estimation’’ at Bernoulli-IMS 11th World Congress in Probability and Statistics, Aug 12-16, 2024. Stay tuned for further details!
January 2024, our paper on ‘‘Distributionally Robust Quickest Detection’’ is accepted by AISTATS 2024.
December 2023, I am organizing a session on ‘‘Recent advances in sequential detection’’ at CMStatistics 2023.
December 2023, our paper on ‘‘Robust Detection using Sequential Uncertainty Sets’’ is accepted by ICASSP 2024. Congrats to Yiran!
October 2023, I give a presentation on Sequential Change Detection at Applied Probability and Statistics Seminar Series, Department of Mathematics and Statistics, Hunter College.
October 2023, I will give a presentation on Robust Change Detection at INFORMS Annual Meeting, Session SB31, Statistical Methods for Operations, October 15, 10:45 AM - 12:00 PM (third speaker).
September 2023, Invited talk at Allerton Conference, UIUC.
August 2023, Invited Plenary talk at IJCAI 2023 Workshop Bridge-AI: from Climate Change to Health Equity.
July 2023, Talk at Joint Conference on Statistics and Data Science in China.
June 2023, Invited Talk at 10th International Workshop on Applied Probability (IWAP), Greece.
May 2023, one conference paper Improving Adversarial Robustness by Contrastive Guided Diffusion Process and one workshop paper MissDiff: Training Diffusion Models on Tabular Data with Missing Values accepted to ICML 2023. Congrats Yidong!
Spring 2023, I am teaching a new undergraduate course, Mathematical Statistics, and a new graduate course, Advanced Statistics Theory, at CUHK-Shenzhen.
November 2022, Invited talk at Department of Statistics and Data Science, Cornell University.
October 2022, Invited Talk at SINE seminar series, Coordinated Science Laboratory, UIUC.