|COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring.|
Matrix Concentration Inequalities via the Method of Exchangeable Pairs
If you have a question about this talk, please contact Zoubin Ghahramani.
I will present an approach to deriving exponential concentration inequalities and polynomial moment inequalities for the spectral norm of a random matrix based on Stein’s method of exchangeable pairs. Our work extends results by Chatterjee on concentration for scalar random variables to the setting of random matrices. When applied to a sum of independent random matrices, this approach yields matrix generalizations of the classical inequalities due to Hoeffding, Bernstein, Khintchine, and Rosenthal. The same technique delivers bounds for sums of dependent random matrices and more general matrix-valued functions of dependent random variables.
This talk is part of the Machine Learning @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
Other listsArmourers and Brasiers Cambridge Forum Annual Disability Lecture Festival of Ideas
Other talksIn vivo models Genomic technologies The virtual fibring theorem for 3-manifolds Intellectual Societies: Intimacy and Knowledge in the 19th Century Bayesian Machine Learning for Controlling Autonomous Systems Regulatory B cells: Novel faces and mechanisms