Bradley Lecture 2024

The University of Georgia, Department of Statistics and The Statistics Club jointly present the 2024 Bradley Lecture on Friday, April 19, 2024 at 4:00pm ET with featured speaker Dr. Dipak Kumar Dey (wiki-site), Board of Trustees Distinguished Professor in the Department of Statistics at the University of Connecticut.

Generalized Variable Selection Algorithms for Gaussian Process Models
With the rapid development of modern technology, massive amounts of data with complex pattern is generated. Gaussian process models that can easily fit the nonlinearity in data become more and more popular nowadays. It is often the case that in some data, only a few features are important or active. However, unlike classical linear models, it is challenging to identify active variables in Gaussian process models. One of the most commonly used methods for variable selection in Gaussian process models is automatic relevance determination, which is known to be open ended. There is no rule of thumb to determine the threshold for dropping features, which makes the variable selection in Gaussian process models ambiguous. In this work, we propose two variable selection algorithms for Gaussian process models, which use the artificial nuisance columns as baseline for identifying the active features. Moreover, the proposed methods work for both regression and classification problems. The algorithms are demonstrated using comprehensive simulation experiments and an application to multi-subject electroencephalography (EEG) data that studies alcoholic levels of experimental subjects.

Dipak Kumar Dey is an Indian-American statistician best known for his work on Bayesian methodologies. He is currently the Board of Trustees Distinguished Professor in the Department of Statistics at the University of Connecticut. Dr. Dey has an international reputation as a statistician as well as a data scientist. Since he earned a Ph.D. degree in statistics from Purdue University in 1980, Dr. Dey has made tremendous contributions to the development of modern statistics, especially in Bayesian analysis, decision science and model selection. Dr. Dey has published more than 10 books and edited volumes, and over 260 research articles in peer-refereed national and international journals. In addition, the statistical methodologies that he has developed has found wide applications in a plethora of interdisciplinary and applied fields, such as biometry and bioinformatics, genetics, econometrics, environmental science, and social science. Dr. Dey has supervised 40 Ph.D. students, and presented more than 200 professional talks in colloquia, seminars and conferences all over the world. Dr. Dey is an elected fellow of the American Association for the Advancement of Science, the American Statistical Association, the Institute of Mathematical Statistics, the International Society for Bayesian Analysis and the International Statistical Institute.  University of Connecticut Biography

Taking place at 145 Brooks Hall / Founders Memorial Garden