Seminars and Colloquia by Series

Fractionally modulated discrete Carleson's Theorem and pointwise Ergodic Theorems along certain curves

Series
Analysis Seminar
Time
Wednesday, February 12, 2025 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Anastasios FragkosGeorgia Institute of Technology

For \( c\in(1,2) \) we consider the following operators
\[
\mathcal{C}_{c}f(x) \colon = \sup_{\lambda \in [-1/2,1/2)}
\bigg| \sum_{n \neq 0} f(x-n) \frac{e^{2\pi i\lambda \lfloor |n|^{c} \rfloor}}{n} \bigg|\text{,}
\]
\[
\mathcal{C}^{\mathsf{sgn}}_{c}f(x) \colon = \sup_{\lambda \in [-1/2,1/2)}
\bigg| \sum_{n \neq 0} f(x-n) \frac{e^{2\pi i\lambda \mathsf{sign}(n) \lfloor |n|^{c} \rfloor}}{n} \bigg| \text{,}
\]
and prove that both extend boundedly on \( \ell^p(\mathbb{Z}) \), \( p\in(1,\infty) \). 

The second main result is establishing almost everywhere pointwise convergence for the following ergodic averages
\[
A_Nf(x)\colon =\frac{1}{N}\sum_{n=1}^N f(T^n S^{\lfloor n^c\rfloor} x) \text{,}
\]
where $T,S\colon X\to X$ are commuting measure-preserving transformations on a  $\sigma$-finite measure space $(X,\mu)$, and $f\in L_{\mu}^p(X), p\in(1,\infty)$. 

The point of departure for both proofs is the study of exponential sums with phases  $\xi_2 \lfloor |n^c|\rfloor+ \xi_1n$ through the use of a simple variant of the circle method.

This talk is based on joint work with Leonidas Daskalakis.
 

What the Hecke is the BMW Algebra?

Series
Geometry Topology Student Seminar
Time
Wednesday, February 12, 2025 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Jake GuyneeGeorgia Tech

The Jones polynomial was first defined by Vaughan Jones as a "trace function" on an algebra discovered via operator algebras. It was discovered that the polynomial satisfies certain skein relations. The HOMFLY polynomial was discovered through both skein relations and a "lift" of the trace function on the Jones algebra to the "Hecke algebra". Another 2-variable polynomial called the Kauffman polynomial was discovered purely via skein relations. In this talk, we discuss how the process started by Jones was reversed for this polynomial. More precisely, we will show how Birman, Wenzl, and Murakami constructed the BMW algebra and a trace function that yields the Kauffman polynomial. We will discuss the significance of the Kauffman polynomial as well as some relationships between the BMW, Hecke, and Jones algebras.

Blow-up of the Modified Benjamin-Ono Equation

Series
PDE Seminar
Time
Tuesday, February 11, 2025 - 15:30 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Justin HolmerBrown University

We prove that negative energy solutions to the modified Benjamin-Ono (mBO) equation, which is L^2 critical, with mass slightly above the ground state mass, blow-up in finite or infinite time.   These blow-up solutions lie adjacent to those constructed by Martel & Pilod (2017) that have mass exactly equal to the ground state mass.  The solutions that we construct, with mass slightly above the ground state mass, are numerically observable and expected to be stable.  This is joint work with Svetlana Roudenko and Kai Yang.

Branched covers and SU(2)-representations

Series
Geometry Topology Seminar
Time
Monday, February 10, 2025 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Juanita Pinzon-CaicedoNotre Dame

The fundamental group is one of the most powerful invariants to distinguish closed three-manifolds, and the existence of non-trivial homomorphisms $\pi_1(M)\to SU(2)$ is a great way of measuring the non-triviality of a three-manifold $M$. It is known that if an integer homology 3-sphere is either Seifert fibered or toroidal, then irreducible representations do exist. In contrast, the existence of SU(2)-representations for hyperbolic homology spheres has not been completely established. With this as motivation, I will talk about partial progress made in the case of hyperbolic homology spheres realized as branched covers. This is joint work with Sudipta Ghosh and Zhenkun Li.

Introduction to reservoir computing

Series
Applied and Computational Mathematics Seminar
Time
Monday, February 10, 2025 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Yunho KimUNIST, Korea

Reservoir computing is a branch of neuromorphic computing, which is usually realized in the form of ESNs (Echo State Networks). In this talk, I will present some fundamentals of reservoir computing from both the mathematical and the computational points of view. While reservoir computing was designed for sequential/time-series data, we recently observed its great performances in dealing with static image data once the reservoir is set to process certain image features, not the images themselves. Hence, I will discuss possible applications and open questions in reservoir computing.

Flat families of matrix Hessenberg schemes over the minimal sheet

Series
Algebra Seminar
Time
Monday, February 10, 2025 - 13:00 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Rebecca GoldinGeorge Mason University

Please Note: There will be a pre-seminar at 10:55 am in Skiles 005.

The flag variety G/B plays an outsized role in representation theory, combinatorics, geometry and algebra. Hessenberg varieties form a special class of subvarietes of the flag variety, arising in diverse contexts. The cohomology ring of a semisimple Hessenberg variety is recognized to be a representation of an associated finite group, and is related to the expansion of some special polynomials in terms of other well-known polynomial bases. These varieties may have pathological behavior, and their basic properties have been characterized only in restricted cases. Matrix Hessenberg schemes in type A consist of a lift of these varieties to G = Gl(n, C), where we can use the coordinate ring of matrices to study them.

In this talk, we present a full characterization of matrix Hessenberg schemes over the minimal sheet of Lie(G) in type A. We show that each semisimple matrix Hessenberg scheme lies in a flat family with a nilpotent matrix Hessenberg scheme, which in turn allows us to study their geometric properties. We describe the schemes fully in terms of Schubert varieties and opposite Schubert varieties, both well-known subvarieties of G/B. More subtly we characterize combinatorially which matrix Hessenberg schemes are reduced. These results are joint with Martha Precup at Washington University, St. Louis. 

Georgia Scientific Computing Symposium

Series
Applied and Computational Mathematics Seminar
Time
Saturday, February 8, 2025 - 08:45 for 8 hours (full day)
Location
Clough 144
Speaker

The Georgia Scientific Computing Symposium (GSCS) is a forum for professors, postdocs, graduate students and other researchers in Georgia to meet in an informal setting, to exchange ideas, and to highlight local scientific computing research. Established in 2009, this annual symposium welcomes participants from the broader research community. The event features a day-long program of invited talks, lightning presentations and ample opportunities for networking and collaboration.  Please check this year's information at https://wliao60.math.gatech.edu/2025GSCS.html

The Allen-Cahn equation with weakly critical initial datum

Series
CDSNS Colloquium
Time
Friday, February 7, 2025 - 15:30 for 1 hour (actually 50 minutes)
Location
Skiles 314
Speaker
Tommaso RosatiU Warwick

Please Note: Zoom link: https://gatech.zoom.us/j/91390791493?pwd=QnpaWHNEOHZTVXlZSXFkYTJ0b0Q0UT09

Motivated by question in the dynamics of phase fields, we study the Allen-Cahn equation in dimension 2 with white noise initial datum. We prove the appearance of a universal initial condition for mean curvature flow in a small noise scaling. We also obtain a weak coupling limit when the noise is not tuned down: the effective variance that appears can be described as the solution to an ODE. I will discuss ongoing applications in the perturbative study of other critical SPDEs. Joint works with Simon Gabriel, Martin Hairer, Khoa Lê and Nikos Zygouras.

High-dimensional tic-tac-toe: how big are the Hales–Jewett numbers?

Series
Combinatorics Seminar
Time
Friday, February 7, 2025 - 15:15 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Misha LavrovKennesaw State University

The Hales--Jewett theorem, one of the fundamental results of Ramsey theory, guarantees that when an $n$-dimensional $t \times t \times \dots \times t$ grid is colored with $r$ colors, if $n$ is sufficiently large depending on $r$ and $t$, then the grid contains a line of $t$ collinear points of the same color (possibly with some further restrictions on the line). If you know a second fact about the Hales--Jewett theorem, it is probably that the upper bounds on $n$ grow incredibly quickly (even after tremendous improvement from Shelah in 1988).

In this talk, we will survey the general upper bounds on the Hales--Jewett numbers and move on to results for specific values of $r$ and $t$. We show an upper bound of ``only'' about $10^{11}$ on $n$ when $r=2$ and $t=4$, and discuss the challenges and open questions in extending this to larger cases.

Adaptive density estimation under low-rank constraints

Series
Stochastics Seminar
Time
Thursday, February 6, 2025 - 15:30 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Olga KloppESSEC Business School and CREST

In this talk, we address the challenge of bivariate probability density estimation under low-rank constraints for both discrete and continuous distributions. For discrete distributions, we model the target as a low-rank probability matrix. In the continuous case, we assume the density function is Lipschitz continuous over an unknown compact rectangular support and can be decomposed into a sum of K separable components, each represented as a product of two one-dimensional functions. We introduce an estimator that leverages these low-rank constraints, achieving significantly improved convergence rates. Specifically, for continuous distributions, our estimator converges in total variation at the one-dimensional rate of (K/n)^{1/3} up to logarithmic factors, while adapting to both the unknown support and the unknown number of separable components. We also derive lower bounds for both discrete and continuous cases, demonstrating that our estimators achieve minimax optimal convergence rates within logarithmic factors. Furthermore, we introduce efficient algorithms for the practical computation of these estimators.

Pages