USF Home > College of Arts and Sciences > Department of Mathematics & Statistics

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

TBA

Steve Wang

Carleton University

Ottawa, Ontario, Canada

3:00pm-4:00pm

CMC 130

Xiang-dong Hou

**Abstract**

TBA

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

TBA

Greg Knese

Washington University

3:00pm-4:00pm

CMC 130

Dima Khavinson

**Abstract**

TBA

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

TBA

William Bogley

Oregon State University

TBA

TBA

Dmytro Savchuk

**Abstract**

TBA

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

TBA

Mark Mineev-Weinstein

New Mexico Consortium

US & International Institute of Physics, Brazil

4:00pm-5:00pm

CMC 130

Razvan Teodorescu

**Abstract**

TBA

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

TBA

Sam Nelson

Claremont McKenna College, CA

3:00pm-4:00pm

CMC 130

Mohamed Elhamdadi

**Abstract**

TBA

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

A geometric approach to understanding neural codes in recurrent networks

Carina Curto

Pennsylvania State University

2:00pm-3:00pm

CMC 108

Nataša Jonoska

**Abstract**

Synapses in many cortical areas of the brain are dominated by local, recurrent connections. It has long been suggested, therefore, that cortical networks may serve to restore a noisy or incomplete signal by evolving it towards a stored pattern of activity. These “preferred” activity patterns are constrained by the network's connections, and are typically modeled as stable fixed points of the dynamics. In this talk I will briefly review the permitted and forbidden sets model for cortical networks, first introduced by Hahnloser et al. (Nature, 2000), and then present some recent results that provide a geometric handle on permitted sets. Specifically, I will show how questions about fixed points can be translated to questions in classical distance geometry. Finally, I will use the geometric description of fixed points to show that these networks can perform error correction and pattern completion for a wide range of connectivities.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Categorification in Applied Mathematics

Robert Ghrist

University of Pennsylvania

1:30pm-2:30pm

CMC 130

Dmytro Savchuk

**Abstract**

One of the organizing principles in Mathematics is that of categorification — the systematic lifting of numerical equalities to isomorphisms of higher algebraic objects: e.g., much of algebraic topology consists of categorification of numerical counts. This talk will be a gentle survey of several ways in which categorification lurks in applied mathematics, with classical and modern ideas alike having enrichments to algebraic structures that reveal richer relationships than numerical equality.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

A topological approach for investigating the intrinsic structure of neural activity

Vladimir Itskov

Pennsylvania State University

3:00pm-4:00pm

CMC 130

Nataša Jonoska

**Abstract**

Experimental neuroscience is achieving rapid progress in the ability to collect neural activity and connectivity data. This holds promise to directly test many theoretical ideas, and thus advance our understanding of “how the brain works”. Detecting meaningful structure in this data is challenging because of unknown nonlinearities, where measured quantities are related to more “fundamental” variables by an unknown nonlinear transformation.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Orthogonal polynomials for normal matrix models with discrete rotational symmetries

Ferenc Balogh

Concordia University

Canada

4:00pm-5:00pm

CMC 130

Razvan Teodorescu

**Abstract**

Statistical observables of random unitary invariant normal matrix models can be expressed in terms of the joint probability distribution of the matrix eigenvalues. This leads to a logarithmic Coulomb gas model in which the eigenvalues are thought of as charged particles in the complex plane under the influence of an external potential.

The joint density of the eigenvalues can be written as a determinant with entries assembled from planar orthogonal polynomials associated to the given background potential. As the matrix size goes to infinity, the asymptotics of eigenvalue statistics, up to leading term, are encoded into the equilibrium measure, the solution of the continuum limit of the Coulomb gas variational problem. To obtain more refined scaling limits, one needs to study the strong asymptotics of the corresponding orthogonal polynomials. After a brief review of the known results to date, I will introduce a special one-parameter normal matrix model with a discrete rotational symmetry for which the equilibrium measure can be found explicitly for all values of the parameter, including a critical value where a non-trivial topological transition of the support is observed. It will be shown how the corresponding orthogonal polynomials can be analyzed using nonlinear steepest descent techniques, based on a trick of writing two-dimensional orthogonality relations in terms of contour integrals, leading to a Riemann-Hilbert problem. In particular, The results confirm a conjectured relation between the limiting zero distribution of the orthogonal polynomials and the equilibrium measure via a balayage procedure.

The talk is based on joint works with T. Grava and D. Merzi.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Why the oracle may not exist: ergodic families of Jacobi matrices, absolute continuity without almost periodicity

Alexander Volberg

Michigan State University

3:00pm-4:00pm

CMC 130

Arthur Danielyan

**Abstract**

We will explain the recent solution of Kotani's problem pertinent to the existence/non-existence of “oracle” (almost periodicity) for the ergodic families of Jacobi matrices (discrete Schröedinger operators). Kotani suggested that such families are subject to the following implication: if family has a non-trivial absolutely continuous spectrum (this happens almost surely) then almost surely it consists of almost periodic matrices (hence the possibility to predict the future by the past). Kotani proved an important positive result of this sort. Recently independently Artur Avila and Peter Yuditskii—myself disproved this conjecture of Kotani (by two different approaches). We will show the hidden singularity that defines when such Kotani's oracle exists or not.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Conformal mapping and random Laplacian growth

Alan Sola

University of Cambridge

3:35pm-4:45pm

CMC 109

Wen-Xiu Ma

**Abstract**

In my talk, I will survey recent advances in the conformal mapping approach to Laplacian random growth in the plane, where aggregating particles are represented by simple conformal maps, and growth of aggregates is encoded through composition of random copies of such maps.

The random planar sets one obtains in this way exhibit rich and fascinating structures, but the analysis of the large-scale geometry and microscopic features of these so-called clusters present a formidable challenge to mathematicians, with many basic questions remaining wide open.

Computer simulations will be used to illustrate the results obtained, and also to formulate a number of open problems.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Asymptotic regimes of complex orthogonal polynomials with varying quartic weight: Global behaviour critical points

Alexander Tovbis

University of Central Florida

3:00pm-4:00pm

CMC 130

Seung-Yeop Lee

**Abstract**

We study the asymptotics of recurrence coefficients for monic orthogonal polynomials \(\pi_n(z)\) with the quartic exponential weight \(\exp\left[-N\left(\frac{z^2}2+\frac{tz^4}4\right)\right]\), where \(t\in\mathbb{C}\) and \(N\in\mathbb{N}\), \(N\to\infty\). Our goal is to describe the regions of different asymptotic behaviour globally in \(t\in\mathbb{C}\) as well as behaviour near the critical points.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

The Homotopy Analysis Method for Differential Equations, and Hierarchies of Integrable Systems

Mathew Baxter

University of Central Florida

2:00pm-3:00pm

CMC 130

Wen-Xiu Ma

**Abstract**

The Homotopy Analysis Method is an innovative new way (Liao, 1992) to get analytical solutions to nonlinear dierential equations. We begin with a brief introduction to the concept of homotopy from topology. From there, the Homotopy Analysis Method is discussed in detail. We describe how the idea of homotopy is applied to introduce a parameter into ordinary/partial differential equations that do not have one to begin with. The homotopy between a linear operator and a nonlinear operator allows us to use perturbation on this parameter to obtain analytical solutions to these equations with small error. The idea is applied to an equation governing the nonlinear evolution of a vector potential of an electromagnetic pulse propagating in an arbitrary pair plasma with temperature asymmetry. We also look at the Hasegawa-Mima equation, a very difficult PDE that governs the electric potential due to a drift wave in a plasma. Future work is discussed.

After this, hierarchies of integrable systems will be discussed. In particular we look at the integrability of the Zakharov-Ito hierarchy due to its zero-curvature representation. We find the bi-Hamiltonian structure, and show the Hamiltonians are involute in pairs under a well-defined Poisson bracket, implying that the equations in our hierarchy are Liouville integrable. Future work is discussed.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Augmented eigenfunctions: a new spectral object appearing in the integral representation of the solution of linear initial-boundary value problems

David Smith

University of Cincinnati

3:45pm-4:45pm

CMC 109

Wen-Xiu Ma

**Abstract**

We study initial-boundary value problems for linear, constant-coefficient partial differential equations of arbitrary order, on a finite or semi-infinite domain, with arbitrary boundary conditions. It has been shown that the recent Unified Transform Method of Fokas can be used to solve all such classically well-posed problems. The solution thus obtained is expressed as an integral, which represents a new kind of spectral transform. We compare the new method, and its solution representation, with classical Fourier transform techniques, and their resulting solution representations. In doing so, we discover a new species of spectral object, encoded by the spectral transforms of the Unified Method.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Quantum invariants of knots, hyperbolic geometry and \(q\)-series

Mustafa Hajij

Louisiana State University

4:00pm-5:00pm

CMC 130

Wen-Xiu Ma

**Abstract**

The discovery of the Jones polynomial lead to a vast family of invariants called the quantum invariants. Quantum invariants deeply connect many domains of mathematics such as quantum groups, hyperbolic geometry, knot theory and number theory. In this talk I will talk about quantum invariants and some of their connections with the geometry of the knot complement. Furthermore, I will describe some recent connections with number theory.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Random Matrices and Potential Theory

Thomas Bloom

University of Toronto

3:00pm-4:00pm

CMC 130

Vilmos Totik

**Abstract**

I will introduce the Gaussian Unitary Ensemble and other unitary ensembles of random matrices.

I will discuss aspects of these ensembles which can be studied and established via potential theory.

Specifically,I will deal with the convergence of the empirical measure of the eigenvalues and large deviation principles.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Algebraic Geometry and Approximation Theory

Michael DiPasquale

University of Illinois at Urbana-Champaign

4:30pm-5:30pm

CHE 302

Wen-Xiu Ma

**Abstract**

Piecewise polynomial functions, also known as splines, are a cornerstone of approximation theory today. A question of fundamental interest in spline theory is to determine the dimension of (and a basis for) the vector space of splines of degree at most \(\mathbf{d}\) over a polytopal complex. We give a survey of some ways in which this question may be approached from the perspective of commutative algebra and algebraic geometry. Key players in this story are the Hilbert function, the Hilbert polynomial, and Castelnuovo-Mumford regularity.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

\(q\)-polynomial invariant of rooted trees; state sum invariant of graphs

Jozef Przytycki

George Washington University

3:30pm-4:30pm

ISA 3050

Mohamed Elhamdadi

**Abstract**

We describe in this note a new invariant of rooted trees and following up state sum invariant of pointed graphs.

We argue that the invariant is interesting on it own, and that it has connections to knot theory and homological algebra. Another reason that we propose this invariant is that we deal here with an elementary, interesting an new mathematics and after the Colloquium everybody can take part in developing the topic inventing new results and connections to other disciplines of mathematics (and likely statistical mechanics and combinatorial biology).

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Pattern formation and bifurcations in reaction-diffusion-advection ecological models

Junping Shi

College of William and Mary

Williamsburg, VA

3:00pm-4:00pm

CMC 130

Yuncheng You

**Abstract**

Spatial-temporal patterns appear often in historical ecosystem data, and the cause of the patterns can be attributed to various internal or external forces. We demonstrate that in spatial ecological models, spatial-temporal patterns can arise as a result of self-organization of the ecosystem. By using bifurcation theory, we show that the spatial-temporal patterns are generated with the effect of diffusion, advection, chemotaxis or time delay.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Planar functions and their importance in cryptography

Stephen M. Gagola

University of Witwatersrand

2:00pm-3:00pm

CMC 130

Brian Curtin

**Abstract**

Planar functions were first introduced by Dembowski and Ostrom. Since 1991 such functions have attracted interest in cryptography as fuctions with optimal resistance to differential cryptanalysis. They were first used in this way by Nyberg where they were given another name “perfect nonlinear” which describes their important cryptographic property of being as far from linear as possible.

Now planar functions have applications in classical cryptographic systems, quantum cryptographic systems, wireless communication, and coding theory. Commutative semifields are equivalent to those planar functions that are known as Dembowski-Ostrom polynomials (DO polynomials). Here I will introduce how Joanne Hall and I have developed methods of constructing families of planar functions and commutative semifields of order \(p^{2r}\) for any odd prime \(p\) and any positive integer \(r\). These families yield a more general construction which includes some other families of known planar functions while at the same time creates new classes of planar functions. Subsequently these were used to construct mutually unbiased bases, a structure of importance in quantum information theory.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Combining Riesz bases

Shahaf Nitzan

Kent State University

4:00pm-5:00pm

CMC 130

Catherine Bénéteau

**Abstract**

Orthonormal bases (ONB) are used throughout mathematics and its applications. However, in many settings such bases are not easy to come by. For example, it is known that even the union of as few as two intervals may not admit an ONB of exponentials. In cases where there is no ONB, the next best option is a Riesz basis (i.e. the image of an ONB under a bounded invertible operator).

In this talk I will discuss the following question: Does every finite union of rectangles in \(R^d\), with edges parallel to the axes, admit a Riesz basis of exponentials? In particular, does every finite union of intervals in \(R\) admit such a basis?

*This is joint work with Gady Kozma.*

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Mathematics in Cryptography: Today's Applications and Tomorrow's Foundations

Rainer Steinwandt

Florida Atlantic University

3:00pm-4:00pm

CMC 130

Brian Curtin

**Abstract**

The first part of the talk will show how mathematical techniques are applied in the design of modern cryptographic protocols. Taking the task of establishing a secret key among \(n\ge2\) users over an insecure network as example, we discuss how computational assumptions enable the derivation of an efficient solution with strong provable guarantees.

Regrettably, some of the most common assumptions needed for today's cryptographic solutions are no longer justifiable in a so-called post-quantum scenario. In particular, popular constructions involving elliptic curves are not available in this setting. Post-quantum cryptography is of interest when cryptographic solutions are expected to guarantee security for many years. The cryptographic community is currently trying to identify mathematical platforms for efficient post-quantum solutions of basic cryptographic tasks like public-key encryption or digital signatures. The second part of the talk will discuss some of the current approaches, including in particular attempts that invoke tools from group theory.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

Self-similar groups and expanding graphs

Ievgen Bondarenko

Kyiv National University

3:00pm-4:00pm

CMC 130

Dmytro Savchuk

**Abstract**

Expanding graphs are highly connected sparse graphs which have numerous applications in mathematics and theoretical computer science. Reingold, Vadhan, and Wigderson (2002) discovered a simple combinatorial construction of expanding graphs. This construction was based on the new operation on regular graphs — the zig-zag product, which is closely related to the replacement product of graphs.

In this talk I will describe how to model iterated replacement product and zig-zag product of graphs by finite automata. Also I will explain a simple construction of self-similar groups whose action graphs produce a family of expanding graphs.

**Title**

**Speaker**

**Time**

**Place**

**Sponsor**

New Challenges in Cryptography — Securing the Cloud and Usersâ€™ Devices

Feng-Hao Liu

University of Maryland

3:00pm-4:00pm

CMC 130

Brian Curtin

**Abstract**

In recent years, we have seen the emergence of cloud computing, where a service provider (the cloud) offers storage and/or computation to individuals, and those individuals can later access their data from various devices. For example, users store documents on Dropbox and later retrieve them via smartphones, tablets, or laptops. This scenario has numerous advantages: it makes it more convenient for individuals to access and share their data, and it can amortize the cost of maintaining a large storage infrastructure. However, as the data may often contain personal or sensitive information, security concerns, such as information leakage, integrity breach, etc., have been a major barrier for individuals, businesses, and organizations in fully adopting the new computing paradigm. To achieve the full power of such paradigm, we must tackle these challenges.

I will talk about my research that explores new cryptographic techniques for ensuring security in the above scenario. I consider emerging threats on two fronts: the remote cloud, and users local devices. With respect to the cloud, I seek techniques to ensure that a compromised provider cannot access users personal information, or return a wrong answer to a request for some computation. With respect to users local devices, one critical security issue I have explored is defending against non-traditional physical attacks implemented by side-channel leakage and/or malicious tampering. To guarantee security, we not only need to introduce new models and definitions of security, but must also develop new algorithmic and analytical techniques to defend against these new classes of attacks.