Robert Tibshirani selected as the 2025 Myles Hollander Distinguished Lecturer
The Department of Statistics at Florida State University welcomed Robert Tibshirani, Professor of Statistics and Professor of Biomedical Data Science at Stanford University, as the 2025 Myles Hollander Distinguished Lecturer.
Tibshirani presented "Univariate-Guided Sparse Regression" on Friday, October 17th, 2025. The recording is available on YouTube.
Lecture Abstract
In this talk I introduce "UniLasso", a novel statistical method for sparse regression. This two-stage approach preserves the signs of the univariate coefficients and leverages their magnitude. Both of these properties are attractive for stability and interpretation of the model. Through comprehensive simulations and applications to real-world datasets, we demonstrate that UniLasso outperforms Lasso in various settings, particularly in terms of sparsity and model interpretability. We prove asymptotic support recovery and mean-squared error consistency under a set of conditions different from the well known irrepresentability conditions for the Lasso. Extensions to generalized linear models (GLMs) and Cox regression are also discussed. A special case of lasso, "uniReg" is an interesting competitor to good ol' least squares regression (Legendre, 1805).
This is joint work with Sourav Chatterjee and Trevor Hastie.
About the Speaker
Robert Tibshirani received a B.S. in statistics and computer science from the University of Waterloo (1979), M.S. in statistics from the University of Toronto (1980), and the Ph.D. in statistics from Stanford University (1984). He has maintained appointments in statistics and the biomedical sciences throughout his career, beginning with his first faculty appointment at the University of Toronto in 1985 in the Department of Statistics and in the Department of Preventive Medicine and Biostatistics. He moved to Stanford University in 1998 and, in addition to his professorship in the Department of Statistics, was professor in the Departments of Public Health Sciences, Health Research and Policy, and, currently, Biomedical Data Science.
Tibshirani’s research contributions provide novel, effective methods that have significantly shaped modern statistical theory and practice. He introduced the Lasso (1996), which is fundamental in high dimensional statistics. He is co-author of five influential books, including Generalized Additive Models (1990), An Introduction to the Bootstrap (1993), and The Elements of Statistical Learning (2001). Recent awards include the COPSS Distinguished Achievement Award and Lectureship, the WNAR Outstanding Impact Award, and the ISI Founders Statistics Prize. He is an elected Fellow of the American Statistical Association, the Institute of Mathematical Statistics, the Royal Society of Canada, and the Royal Society. He is an elected member of the National Academy of Sciences.
The Myles Hollander Distinguished Lectureship is sponsored by:
