Return to Colloquia & Seminar listing
Sparse and Smooth Function Estimation in Reproducing Kernel Hilbert Spaces
Mathematics of Data & DecisionsSpeaker: | Hao Helen Zhang, U. Arizona |
Related Webpage: | http://www.math.arizona.edu/~hzhang |
Location: | Zoom |
Start time: | Tue, Oct 12 2021, 1:10PM |
Curse of dimensionality refers to sparse phenomena of high-dimensional data, and it presents substantial challenges in the theory and computation of nonparametric models. This talk will present a class of regularization operators which enable sparse and smooth estimation of multi-dimensional functions in reproducing kernel Hilbert spaces. The operator leads to a unified framework for model selection to enhance the accuracy and interpretability of a variety of nonparametric models, including generalized additive models, partially linear models, and functional additive models. We discuss the theoretical properties of the estimator and demonstrate its empirical performance in real-world examples.
Dr. Zhang is available for individual or group meetings on Zoom after the talk until 16:00. Contact the seminar organizer
for scheduling.