ASA Connect

Dear Colleagues, 

The Section of Statistical Learning and Data Science is pleased to present the next webinar on unsupervised domain adaption.  Unsupervised domain adaptation concerns how information and knowledge from a source population can be leveraged and transferred to a target population where labels/outcomes are unavailable.  Dr. Zhao from University of Wisconsin-Madison will discuss about estimation and inference in this setting.  Hope to see you there!  

Title:                        Estimation and Inference in Unsupervised Domain Adaptation

Speakers:               Dr. Jiwei Zhao, Associate Professor at the University of Wisconsin-Madison

Date and Time:      December 3, 2025, 2:00 to 3:30 pm Eastern Time

Registration Link:  ASA SLDS Webinar Registration Link [eventbrite.com] 

Abstract:                Unsupervised domain adaptation concerns how information and knowledge from a source population can be leveraged and transferred to a target population where labels/outcomes are unavailable. In this talk, I will present recent results on estimation and inference in this setting, based on a sequence of joint work with my collaborators. I will begin with the case without any distributional shift between the source and target populations, where the objective is to make safe and optimal use of the unlabeled data. The connection to the prediction-powered-inference (PPI) literature will also be discussed. I will then present the covariate-shift and label-shift settings, where the focus is usually on achieving robustness, efficiency, and understanding their tradeoffs. Illustrative examples will be provided to demonstrate the performance of the proposed methods in empirical applications.

Presenter:              Jiwei Zhao is currently an Associate Professor at the University of Wisconsin-Madison, affiliated with the Departments of Statistics and of Biostatistics & Medical Informatics. His research interests include semiparametric statistics, the tradeoff between efficiency and robustness, domain adaptation and transfer learning, missing data analysis and causal inference. His work has been published in top-tier statistical journals as well as in leading machine learning conferences. His research has been consistently supported by the US National Science Foundation and the National Institutes of Health. Jiwei is now Associate Editor (or Action Editor) for Annals of Applied Statistics, JRSS Series A, Scandinavian Journal of Statistics, Journal of Nonparametric Statistics, Transactions on Machine Learning Research, and has been the Area Chair for the annual International Conference on Artificial Intelligence and Statistics (AISTATS) since 2024.