Seminars and Colloquia by Series

Thursday, April 2, 2015 - 15:05 , Location: Skiles 006 , Michael Anshelevich , Texas A&M , Organizer: Ionel Popescu
I will discuss the limit theorems for composition of analytic functions on the upper-half-plane, and the analogies and differences with the limit theorems for sums of independent random variables. The analogies are enhanced by recalling that the probabilistic limit theorems are really results about convolution of probability measures, and by introducing a new binary operation on probability measures, the monotone convolution.This is joint work with John D. Williams.
Thursday, February 26, 2015 - 15:05 , Location: Skiles 006 , Yuri Bakhtin , Courant Institute of Mathematical Sciences, New York University , Organizer: Christian Houdre
Ergodic theory of randomly forced space-time homogeneous Burgers equation in noncompact setting has been developed in a recent paper by Eric Cator , Kostya Khanin, and myself. The analysis is based on first passage percolation methods that allow to study coalescing one-sided action minimizers and construct the global solution via Busemann functions. i will talk about this theory and its extension to the case of space-continuous kick forcing. In this setting, the minimizers do not coalesce, so for the ergodic program to go through, one must use new soft results on their behavior to define generalized Busemann functions along appropriate subsequences.
Thursday, February 12, 2015 - 15:05 , Location: Skiles 006 , Piotr Nayar , IMA, Minneapolis , , Organizer: Christian Houdre
We define the class of ultra sub-Gaussian random vectors and derive optimal comparison of even moments of linear combinations of such vectors in the case of the Euclidean norm. In particular, we get optimal constants in the classical Khinchine inequality. This is a joint work with Krzysztof Oleszkiewicz.
Friday, October 3, 2014 - 14:05 , Location: Skiles 006 , Victor-Emmanuel Brunel , CREST and Yale University , Organizer: Karim Lounici
In this talk we will consider a finite sample of i.i.d. random variables which are uniformly distributed in some convex body in R^d. We will propose several estimators of the support, depending on the information that is available about this set: for instance, it may be a polytope, with known or unknown number of vertices. These estimators will be studied in a minimax setup, and minimax rates of convergence will be given.
Thursday, September 25, 2014 - 15:05 , Location: Skiles 006 , Ionel Popescu , Georgia Tech , Organizer: Ionel Popescu
 The CLT for free random variables was settled by Voiculescu very early in this work on free probability.   He used this in turn to prove his main result on aymptotic freeness of independent random matrices.   On the other hand, in random matrices, fluctuations can be understood as a second order phenomena.  This notion of fluctuations has a conterpart in free probability which is called freenes of second order.  I will explain what this is and how one can prove a free CLT result in this context.   It is also interesting to point out that this is a nontrivial calculation which begs the same question in the classical context and I will comment on that.   
Thursday, September 11, 2014 - 15:05 , Location: Skiles 006 , Vladimir Koltchinskii , School of Mathematics, Georgia Tech , Organizer: Christian Houdre
We will discuss sharp bounds on moments and concentration inequalities for the operator norm of deviations of sample covariance operators from the true covariance operator for i.i.d. Gaussian random variables in a separable Banach space. Based on a joint work with Karim Lounici.
Thursday, September 4, 2014 - 15:05 , Location: Skiles 006 , Christian Houdre , School of Mathematics, Georgia Tech , Organizer: Christian Houdre
Let (X_k)_{k \geq 1} and (Y_k)_{k\geq1} be two independent sequences of independent identically distributed random variables having the same law and taking their values in a finite alphabet \mathcal{A}_m. Let LC_n be the length of the longest common subsequence of the random words X_1\cdots X_n and Y_1\cdots Y_n. Under assumptions on the distribution of X_1, LC_n is shown to satisfy a central limit theorem. This is in contrast to the Bernoulli matching problem or to the random permutations case, where the limiting law is the Tracy-Widom one. (Joint with Umit Islak)
Thursday, April 17, 2014 - 15:05 , Location: Skiles 005 , Lionel Levine , Cornell University , Organizer: Ionel Popescu
A sandpile on a graph is an integer-valued function on the vertices. It evolves according to local moves called topplings. Some sandpiles stabilize after a finite number of topplings, while others topple forever. For any sandpile s_0 if we repeatedly add a grain of sand at an independent random vertex, we eventually reach a sandpile s_\tau that topples forever. Statistical physicists Poghosyan, Poghosyan, Priezzhev and Ruelle conjectured a precise value for the expected amount of sand in this "threshold state" s_\tau in the limit as s_0 goes to negative infinity. I will outline the proof of this conjecture in and explain the big-picture motivation, which is to give more predictive power to the theory of "self-organized criticality".
Thursday, April 10, 2014 - 15:05 , Location: Skiles 005 , Irina Holmes , Louisiana State University , Organizer: Ionel Popescu
In this talk we investigate possible applications of the infinitedimensional Gaussian Radon transform for Banach spaces to machine learning. Specifically, we show that the Gaussian Radon transform offers a valid stochastic interpretation to the ridge regression problem in the case when the reproducing kernel Hilbert space in question is infinite-dimensional. The main idea is to work with stochastic processes defined not on the Hilbert space itself, but on the abstract Wiener space obtained by completing the Hilbert space with respect to a measurable norm.
Thursday, April 3, 2014 - 15:05 , Location: Skiles 005 , Umit Islak , University of Southern California , Organizer: Ionel Popescu
Let $Y$ be a nonnegative random variable with mean $\mu$, and let $Y^s$, defined on the same space as $Y$, have the $Y$ size biased distribution, that is, the distribution characterized by $\mathbb{E}[Yf(Y)]=\mu \mathbb{E}[f(Y^s)]$ for all functions $f$ for which these expectations exist. Under bounded coupling conditions, such as $Y^s-Y \leq C$ for some $C>0$, we show that $Y$ satisfies certain concentration inequalities around $\mu$. Examples will focus on occupancy models with log-concave marginal distributions.