Oct 5
Herman K. van Dijk
Erasmus University Rotterdam
Oct 12
Allen Gorin
U.S. Department of Defense, Fort Meade, MD
Oct 19
Blake McShane
Northwestern University
Oct 26
Carlos Carvalho
University of Texas at Austin
Nov 2
Robert Vanderbei
Princeton University
Nov 9
Shankar Bhamidi
University of North Carolina
Nov 16
Larry Brown
University of Pennsylvania
Nov 30
Fan Li
Duke University
Dec 7
David Madigan
Columbia University
|
Elchanan Mossel
University of California, Berkeley
Wednesday, September 28, 4:30-5:30 PM
F60 Jon M. Huntsman Hall
Statistics – PRiML Joint Seminar
Some Recent Progress in Combinatorial Statistics
Combinatorial statistics deals with estimation of discrete parameters where the goal is to reconstruct the parameters exactly using explicit bounds on the number of samples needed, the running time of the estimation procedure and the estimation accuracy. I will discuss some recent work in this area including estimation of Markov random fields and estimation of "noisy" rankings.
Thursday, September 29, 4:30-5:30 PM
265 Jon M. Huntsman Hall
Statistics – CIS Joint Seminar
Non-linear Invariance and Applications
Over the last decade, non-linear invariance principles for low-influence functions have played a major role in the theory approximation algorithms in computer science and in the theory of voting schemes in theoretical economics. The talk will provide a broad overview of non-linear invariance, Gaussian geometry and their connection to hardness of approximation and social choice theory.
Friday, September 30, 2:00-3:00 PM
100 Towne Building
Statistics – AMCS Joint Seminar
On Reverse Hypercontractive Inequalities
A hyper-contractive inequality for an operator T states that |Tf|q <= |f|p where q > p > 1 for all functions f. Hyper contractive inequalities play a crucial role in analysis in general and indiscrete Fourier analysis in particular. A reverse hyper-contractive inequality for the operator T states that |Tf|q >= |f| p for q < p < 1 (q and p can be negative) and all strictly positive functions f. The first reverse hyper-contractive inequalities were proved by Borell more than 2 decades ago. While these inequalities may look obscure, they have been used for the solution of a number of problems in the last decade. I will survey applications of the inequalities and discuss new results relating reverse hyper-contractive inequalities to hyper-contractive, Log-Sobolev and Poincare inequalities as well as some new applications. This is a joint work with K. Oleszkiewicz (Warsaw) and A. Sen (Cambridge).
Refreshments will be served on 9/28 and 9/29
at 4:00 pm in 440 Huntsman Hall
Check out our website for details about upcoming seminars: http://statistics.wharton.upenn.edu/news/Seminars.cfm |