April 19 SFASA
Seminar
Speaker: Robert Tibshirani, Stanford
University
Rob
Tibshirani is Professor of Health Research and Policy (Biostatistics) and is
Vice-Chair of the Department. He is also Professor of Statistics at
Stanford. He is the inventor of the celebrated lasso technique, and is
co-author of three books. They include one on additive regression with
Trevor Hastie, one on the bootstrap with Bradley Efron, and one on many aspects
of modern machine learning with Trevor Hastie and Jerry Friedman. Rob won
the COPSS Presidents' Award in 1996. For further material see http://en.wikipedia.org/wiki/Robert_Tibshirani
Location: Stanford
University, Li Ka Shing Center (LKSC) Building, 291 Campus Drive, Room: LK 120
Seminar Classroom
Time: Thursday,
April 19th, 4:30 p.m. – 5:45 p.m.
Title: The Lasso: some novel
algorithms and applications Abstract:
I
will discuss some procedures for modeling high-dimensional data, based on L1
(lasso)-style penalties. I will describe pathwise coordinate descent algorithms
for the lasso, which are remarkably fast and facilitate application of the
methods to very large datasets for the first time. I will then give examples of
new applications of L1 penalties to microarray classification, near-monotonic
regression, and sparse principal components.
Directions and maps: Please click on the following link to see a
campus map that shows the location of the LKSC Building on StanfordŐs campus:
http://med.stanford.edu/irt/about/maps-directions/lksc-directions.html
Return to Bay
Area ASA Homepage