When
Where
Speaker: Gautam Dasarathy, School of Electrical, Computer & Energy Engineering, ASU
Title: Label-Efficient Two-Sample Testing
Abstract: Hypothesis testing is a classical statistical inference approach used to determine whether the data at hand supports a specific hypothesis. An important type of hypothesis test is the two-sample test, which evaluates whether two sets of data points or samples are from identical distributions. This framework is widely used in various domains, from clinical trials to online advertisements, where one aims to understand the effectiveness of a treatment or an ad. In this talk, we will explore two-sample testing in a context where an analyst has access to lots of data (features) from the two samples, but determining sample membership (or labels) of these features is costly. We will show how we can combine techniques from (interactive) machine learning and classical statistics to perform effective two-sample testing in this resource-constrained setting while maintaining statistical validity and attaining high testing power. We will also discuss several applications of this framework and some future directions.
Bio: Gautam Dasarathy received a Master’s and Ph.D. in Electrical Engineering from the University of Wisconsin at Madison in 2010 and 2014, respectively. He was then a Post-Doctoral Researcher at Rice University, Houston, TX, and Carnegie Mellon University, Pittsburgh, PA. He is an Associate Professor at the School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe, AZ, USA. His research interests include machine learning, statistics, information processing, and networked systems. His work has received support from funding agencies across diverse domains, including the National Science Foundation (NSF), Defense Advanced Research and Projects Agency (DARPA), Office of Naval Research (ONR), and the National Institutes of Health (NIH). Dr. Dasarathy received the CAREER Award from the NSF in 2021. He also received the Distinguished Alumnus Award from his undergraduate alma mater, VIT University (India), in 2022.