Title: Sparse and Smooth Function Estimation in Reproducing Kernel Hilbert Spaces
Abstract: My talk consists of two parts. The first part will present my research in high-dimensional nonparametric regression for model estimation, variable selection, and inferences. A new class of regularization frameworks is introduced in the context of reproducing kernel Hilbert spaces to achieve simultaneous sparsity and smoothness in multivariate function estimation. Their theoretical properties and empirical performance are discussed. In the second part, I will provide an overview of UA-TRIPODS. Deep-level collaborations among theoretical computer science, mathematics, and statistics are indispensable to build theoretical foundations of data sciences. Our aim is to build a highly collegial research institute that brings three communities together to tackle fundamental problems in data science with the “common language” and “same mindset”. I will share our three-year experiences, highlighting our strategies, pitfalls, as well as achievements in research, education, and outreach.
Bio: Hao Helen Zhang is Professor of Mathematics and Statistics Interdisciplinary Program at the University of Arizona. Her research areas are in statistical machine learning, high dimensional analysis, nonparametric smoothing, and variable selection. She has published more than 80 research articles and a book "Principles and Theory for Data Mining and Machine Learning". Her research has been funded by NSF, NIH, NSA, including the NSF CAREER award and TRIPODS. Dr. Zhang is Editor-in-chief of STAT and Associate Editor of JASA and JRSS-B. She is Fellow of the American Statistical Association, Fellow of the Institute of Mathematical Statistics, and the 2019 IMS Medallion Lecturer.