Header image   header image 2


Organizers at USC Sumter

Rigoberto Flórez

Phone: (803) 938 3886

Leandro Junes

Organizers at USC Lancaster

Shemsi Irene Alhaddad

Jason Holt

Andrew Yingst

Organizers at USC Salkehatchie

Wei-Kai Lai

Fidele Ngwane

Organizers at Benedict College

Alexandru Gabriel Atim

Kwan Lam

Organizers at The Citadel

Upasana Kashyap

Antara Mukherjee

The url will change soon to http://macs.citadel.edu/florez

Fall Meeting at Citadel
Oct 28 2011

Kristopher Liggins

On Generalization of Fibonacci Numbers


Chris Rufty

A Japanese Ladder Game, a simple version

In MAA FOCUS newsmagazine Vol.31, No.3 June/July 2011, Dr. Dougherty and Dr. Vasquez introduced a puzzle, known as Japanese Ladder Game. It is known that every Japanese ladder will provide a 1-1 mapping, and for any sequence at the bottom of the ladder, one can always find a minimum solution. By studying a simpler version, we were able to use our method to prove some of the old results. In this talk, we will introduce the way to play the game, and the basic Math related to this game. Finally, we will give two simple proofs of our theorems.


Ralph Howard

Integral Geometry with Applications to Geometric Inequalities

We outline a few of the basic results in integral geometry, and their application to geometric inequalities such as the Bonnesen inequality (a refinement of the isoperimetric inequality) and Fray-Milnor inequality on the total curvature of knots. If time permits some recent work on knot energy will be given. The talk should be accessible to undergraduates with a knowledge of vector calculus.


Chuck Groetsch

Inverse problems, von Neumann’s theorem, and stable approximate evaluation of unbounded operators

The mathematical framework for many linear inverse problems in the mathematical sciences requires the application of an unbounded operator to a vector which is not in its domain – an impossible task! We present a brief introduction to a general approach, based on a classical theorem of von Neumann, for addressing this difficulty.


Oleg Smirnov

Lie Algebras, Groups Triple Systems: the Truth

In the talk we consider functorial connections between the categories of  the Lie algebras, Lie groups, Lie triple systems, and Symmetric Spaces.

We present a full subcategory of graded Lie algebras which is equivalent to the category of triple systems and discuss a similar connection between Lie groups and symmetric spaces.


Mei Chen

Eigenpairs of Adjacency Matrices of Balanced Signed Graphs

In this talk, we present results on eigenvalues λ  and their associated eigenvectors x of an adjacency matrix A of a balanced signed graph. A graph G =(V,E) consists of a set V of vertices and a set E of edges between two adjoined vertices. A signed graph is a graph for
which each edge is labeled with either + or -. A signed graph is said to be balanced if there
are an even number of negative signs in each cycle (a simple closed path).

Signed graphs were first introduced and studied by F. Harary to handle a problem in social psychology. It was shown by Harary in 1953 that a signed graph is balanced if and only if its vertex set V can be divided into two sets (either of which may be empty), X and Y, so that each edge between the sets is - and each within a set is +. Based on this fundamental theorem for balanced signed graphs, vertices of a balanced signed graph can be labeled in a way so that its adjacency matrix is well structured. Using this special structure, we find exactly all eigenvalues and their associated eigenvectors of the adjacency matrix A of a given balanced signed graph. We will present eigenpairs ( λ, x) of adjacency matrices of three types of balanced signed graphs: (1) graphs that are complete; (2) graphs with t vertices in X or in Y that are not connected; and (3) graphs that are bipartite

Joint work with Spencer P. Hurd of The Citadel.


Previous Meetings abstracts

Fall Meeting at USC Sumter
September 16 2011


Shannon Michaels Areford

A Relation Between Fibonacci Numbers and Lucas Numbers.

In this talk I discuss some elementary properties of Fibonacci numbers. I will prove an identity that shows that the difference of two Fibonacci squares is a Lucas Number.


Thomas L. Fitzkee

Topology Explains Why Automobile Sunshades Fold Oddly.

The article uses topology and abstract algebra to examine "automatic folding" sunshades that coil up when not in use. From the authors' experience, it seems impossible simply to fold such a sunshade in half (i.e. coil it into exactly two loops). The object here is to figure out how many loops can appear in the coil and to understand why.

Kevin Milans

Subtrees with Few Labeled Paths.

Consider a {0,1}-edge-labeling of the complete rooted ternary tree T of depth n. The edge labels along a path from the root to a leaf produce a bitstring of length n; such a bitstring is called a path label. For each complete binary subtree S of depth n, let L(S) be the set of path labels that occur along paths in S. We study the problem of finding a subtree S such that |L(S)| is small. The problem originated from a question in computability theory. This is joint work with Rod Downey, Noam Greenberg, and Carl Jockusch.


Jason Burns

Extending Hadwiger's characterization theorem.

In 1957, Hugo Hadwiger classified the continuous, rigid-motion invariant, finitely additive measures on convex sets in n-dimensional Euclidean space. Let me explain those buzzwords and persuade you this is important, then prove a 'baby' version and persuade you this is true, then talk about extensions to non-Euclidean space and persuade you there is more to be discovered.



Leandro Junes

Convolutions on the Geometry of Fibonacci numbers.

We define a discrete convolution C using Fibonacci numbers that acts on the Hosoya’s triangle. We prove that this convolution give rise to some counting theorems. Those theorems are used to count some words’ patterns in formal languages. In particular, we have found that that those theorems can be used to count the weight of non-decreasing Dyck paths. Work in progress in collaboraton with Rigo Florez.


Abstracts, meeting at Lancaster April 15 2011


Julia Smith (Student at USC Sumter)

Inverse functions and ciphers

A cipher is a method for creating secret messages. The purpose of using a cipher is to exchange information securely. Throughout history, many different methods have been created. I will discuss a cipher that depends on a linear function and the use of its inverse to decode a secret message. Basic modular arithmetic will be used throughout the talk.



Linyuan Lu

A Fractional Analogue of Brooks' Theorem

Abstract: Let Δ be the maximum degree of a connected graph G. Brooks' theorem states that the only connected graphs with chromatic number Δ+1 are complete graphs and odd cycles. Here we proved a fractional version of Brooks' theorem:  we classified all connected graphs G with the fractional chromatic number χf(G) ≥ Δ. (Joint work with Xing Peng)



Naima Naheed

Convex Minorant of the Nonconvex Thomas-Fermi Energy Functional

Mathematically rigorous versions of Thomas-Fermi theory and its generalizations were developed in the 1970s and 1980s by Lieb, Simon, Benilan, Brezis, Gisele and Jerome Goldstein and others. Later Phillippe Benilan, Gisele and Jerome Goldstein (BGG) incorporated Fermi-Amaldi correction into the Thomas-Fermi energy functional. As a result convexity is lost.

The theory which will be presented here includes a convex minorant of the nonconvex Thomas-Fermi energy functional. The corresponding Euler-Lagrange equation will become a nonlinear elliptic system involving measures, which will be solved using the methods of BGG. Then the existence of a ground state Thomas-Femi density will be obtained using, among other things, topological degree theory.



Xing Peng

The minimum number of monochromatic short progressions in Zn

Abstract: For any n≥k≥3, let Mk(n) be the minimum number of monochromatic k-term progressions in any 2-coloring of Zn. We studied asymptotic bounds ofMk(n) for k=3,4 and large n.  For k=3, we show random colorings achieve the minimum number of monochromatic 3-term progressions.  For k=4, we construct a 2-coloring of Zn with few 4-term progressions than a random coloring has. Our upper bound and lower bound for M4(n) improve the previous result given by Wolf on M4(p) for prime p. (Joint work with Linyuan Lu)


Julian Buck

The relationship between Topological Dynamics and C*-Algebras

The classification program for C*-algebras is one of the main current branches of research in abstract functional analysis. C*-algebras that arise by looking at dynamical systems on topological spaces have provided especially good examples for the purpose of classification theory, where one can start with commutative C*-algebras (essentially, function spaces) and construct new noncommutative examples through a universal construction (the crossed product). This produces a sort of 3-tiered system: the base dynamical system, its associated function algebra, and the new crossed product C*-algebra. In this talk I will describe some of the interplay between the dynamical system and its crossed product, and how this is exploited to show the crossed product has a suitably nice structure for the classification program.





Abstracts, meeting at Benedict College Feb 11 2011


Keyona James

African-Americans in Mathematics

In this talk we will discuss contributions made by African-American mathematicians and their involvement in the mathematical sciences. Not much is known, taught, or written about African-American mathematicians. Information lacks on their past and present contributions and on the qualitative and quantitative nature of their existence throughout mathematics.  In this talk we will provide some details about great mathematicians such as Benjamin Banneker, Elbert Frank Cox, Evelyn Boyd Granville, J. Ernest Wilkins, Jr., etc.



Peter Nyikos

Discontinuities and smooth curves in n-space

The following theorem, with various wordings, can be found in a number of calculus texts.

Theorem 1. If the limit of a real-valued function on R2 exists at a point  p, then it will also be the limit along any smooth curve through  p.

This theorem clearly extends to all n2.

The converse is true, but seems to be missing from all the standard calculus textbooks.  In fact, something stronger is true:

Theorem 2.  If  f  is a real-valued function  defined in a deleted neighborhood of  p  in Rm, and the limit of  f  at  p  does not exist, then either:
(1) there is a smooth curve through  p  on which the limit does not exist, or
(2) there are two straight lines through   p on which the limits exist, but are unequal.



László Zsilinszky

On some topological games

The so-called Banach-Mazur game was first introduced and studied in the 1930's by Stanislaw Mazur and Stefan Banach,, who played it on the unit interval. Oxtoby generalized and studied the game in topological spaces, later Choquet  rediscovered it, and introduced a modification, which is now termed the strong Choquet game, to characterize complete metrizability in a metrizable setting. The talk will be an introduction to these games with a review of some interesting related results and examples.


Gurcan Comert

Traffic parameter estimation from probe vehicles at signalized intersections

Instrumented vehicle data (i.e., probe data) is becoming more important for real-time system parameter estimation in transportation networks. Probe data can be tracked anonymously and can report data on their locations, speeds, travel times, and arrival times as they perform their regular trips. This research develops analytical models for the real-time estimation of key traffic parameters (e.g., queue length, delay) at signalized intersections using the fundamental information (i.e., location, count, and time). For a single queue with Poisson arrivals, analytical models are developed to evaluate how error changes in estimation as percentage of probe vehicles in the traffic stream varies. The formulations presented assess the error in estimation for various scenarios of probe vehicle market penetration rates and congestion levels.


Ralph Howard

The Geometry of Mirrors

We use the principle of least action to show why light bounces off a mirror so that the angle of incidence equals  the angle of reflection.  One geometric consequence of this is  that when looking a mirror, one sees a left right reversed version of one's self.  Another is that light bouncing off of a corner will return parallel to its original path.  We also discuss how to construct the angles of a kaleidoscope so as to have non-overlaping images. There will be mirrors and kaleidoscopes for hands on experimenting  


Abstracts, meeting USC Salkehatchie Nov 5 2010


Fidele Ngwane

Computing Integral Closures

Monomial orderings  will be  discussed.  They  are  vital  in our polynomial  computations.   Integral  extensions  will  be   analyzed, in particular, type I integral extensions.  Integral   closures of type I  integral   extensions  have  great applications.  We will  present  a method  for computing integral  closures that is different  from   others.


Balaji Iyangar

Multigrid Methods

Iterative processes for solving the algebraic  equations  arising  from  discretizing  partial differential equations are stalling numerical processes, in which  the  error  has  relatively  small  changes  from one  iteration  to  the  next.  The  computer  grinds  very hard  for  very  small  or  slow  real  physical  effect with the use of too-fine discretization grids. In this case, in large parts of  the computational domain  the meshsize is  much  smaller  than  the  real  scale  of  solution changes.  Such  problems  can  be  overcome  by  the multigrid method, or more generally,  the Multi-Level Adaptive  Technique  (MLAT).  Stalling  numerical processes  are  usually  related  to  the  existence  of several  solution  components  with  different  scales, which conflict with each other. By using interactively several  scales  of discretization,  multigrid  techniques resolve such conflicts, avoiding stalling and also being computationally  efficient.

Antara Mukherjee

Isoperimetric Inequalities Using Varopoulos

I will start by introducing Dehn functions and then show the compu tation of upper bounds of the second order Dehn functions for lattices of three-dimensional geometries, namely Nil and Sol. These upper bounds are obtained by using the Varopoulos transport argument on dual graphs. The idea is to reduce the original isoperimetric problem involving volume of three-dimensional balls and areas of their boundary spheres to a problem involving Varopoulos' notion of volume and boundary of nite domains in dual graphs.

Upasana Kashyap:

A Morita theorem for dual operator algebras.

We consider some new variants of the notion of Morita equivalence appropriate to algebras of Hilbert space operators which are closed in the `weak* -topology' (or equivalently, which are dual spaces and known as dual operator algebras), and we will describe how the earlier theory of strong Morita equivalence due to Blecher, Muhly, and Paulsen, transfers to this `weak*-topology setting'. We will present our main theorem, that two dual operator algebras are weak*-Morita equivalent in our sense if and only if they have equivalent categories of dual operator modules. A key ingredient in the proof of our main theorem is W*-dilation, which connects the non-selfadjoint dual operator algebra with the W*-algebraic.



Abstracts, meeting USC Sumter September 17 2010


Jason Holt:

Non-Random Perturbations of the Anderson Hamiltonian and Cwikel-Lieb-Rozenblum Type Estimates

We will consider the Anderson Hamiltonian H0 = ∆ + V (x; ω) where V is a random potential and \omega belongs to a probability space (ΩF; P). The main object of the present work is the perturbed operator H = H0 -W where W(x)≥ 0 decays at infinity. It is known that the spectrum of H below 0 is discrete consisting only of eigenvalues and that the total number N0(W) of eigenvalues below 0 is a random variable for which P{N0(W) < ∞ } = 1, or P{N0(W) < ∞ } = 0. We develop general conditions on V and W to guarantee P-a.s one case or the other and present several examples demonstrating the borderline decay in W. In particular, it will be shown that if V has a Bernoulli structure, then the borderline between finitely and infinitely many eigenvalues is obtained with a decay in W as O(c0 ln-2 |x| where c0 is a determined positive constant.



Joshua Cooper

Tree reconstruction and a Waring-type problem on partitions

Abstract: The ``line graph'' of a graph G is a new graph L(G) whose vertices are the edges of G, with a new edge in L(G) from e to f if e and f were incident in G. Graham's Tree Reconstruction Conjecture says that, if T is a tree (a connected, acyclic graph), then the sequence of sizes of the iterated line graphs of T uniquely determine T. That is, T can be reconstructed from | L (j) (G)|j=0 , where L (0) (G) = G and L (j+1) (G) = L(L (j) (G)). Call two trees equivalent if they yield the same sequence; we call the resulting equivalence classes ``Graham classes.'' Clearly, the conjecture is equivalent to the statement that the number of Graham classes of n-vertex trees is equal to the number of isomorphism classes of such trees, which is known to be about 2.955765n.

We show that the number of Graham classes is at least superpolynomial in n (namely, exp(c log n 3/2)) by converting the question into the following Waring-type problem on partitions. For a partition λ = {λ_1,...,λ_k} of the integer n and a degree d polynomial f ∈ Z[x], define f(λ) = ∑ k j=1 f(λj). We show that the range of f(λ) over all partitions λ of n grows as Ω(n d-1). The proof employs a well-known family of solutions to the Prouhet-Tarry-Escott problem. Strong evidence suggests the conjecture that the size of the range is actually Θ(nd).

Joint work with Bill Kay of USC



Kwan Lam:

The formation of Turing pattern on networks of complete

In this talk, we will discuss the formation of Turing pattern in networks of homogeneous coupled reactors with a focus on the a specific reaction diffusion model - Lengyel-Epstein kinetics.

Special attention will be paid to the formation of bimodal pattern on the complete graph. We will use it as a building block to construct the linear and circular chains of the complete graphs with bimodal patterns.



Rigoberto Florez

Some open questions

In this talk I am going to discuss some open questions in number theory and combinatorics. If the time allows us, I also go to discuss some potential research problems.







Abstracts, meeting USC Lancaster April 23 2010


Daniel Savu:

On Sparse Approximation in Banach Spaces

The sparse approximation problems ask for complete recovery of functions in a given space that are supported by few of the elements of a system of generators for the space or for approximate recovery that involves a limited number of generators. This is made in regard with redundant systems which offer convenience of representation as well as better rates of approximation. The redundancy raises, in turn, very difficult theoretical problems. We give answers to some of these problems in the very general setting of Banach spaces. The theoretical results complete the previous findings in greedy approximation in this setting and show, for the algorithms considered, the same general recovery properties as the ones known in the particular case of Hilbert spaces. 

 Moreover, we provide a novel idea of improvement of the geometry of the redundant systems by switching to a different setting than the standard Hilbert space. This improvement would translate in better recovery properties as we are able to prove the same efficiency of the greedy approach in the new setting.

Wei-Kai Lai:

The Radon-Nikodym Property for Positive Tensor Products of Banach Lattices

In 1950’s, Grothendieck started the theory of projective and injective tensor products of Banach spaces. From the positivity perspective, Fremlin and Wittstock extended the theory to projective and injective tensor products of Banach lattices in 1972 and 1974 respectively. In this talk, we are going to discuss the Radon-Nikodym property for Fremlin and Wittstock’s versions of tensor product of Banach lattices.

Wei-Tian Li:

Lubell Function and Forbidden Subposets. Part II

In 1928, Sperner proved that the size of a largest antichain in the Boolean lattice Bn is equal to n choose n/2. Since then, the results on largest sizes families not containing some specific posets were discovered gradually. The well-known inequality, the LYM-inequality, was individually used by Lubell, Yamamoto, and Meshalkin to reprove Sperner's Theorem. We will introduce the Lubell function, derived from LYM-inequality, and use it to estimate the maximal sizes of families do not contain some posets P. (This is a joint work with Jerrold R. Griggs and Lincoln Lu.)

Yiting Yang:

On the second order Randic index of trees

Let G be a simple graph. The second Randic index of G is defined as

2R(G)=∑xyz 1/(dxdydz)1/2

where the summation runs over all paths xyz of length two, contained in G. It was first considered by chemists Randic, Kier and Hall in the study of branching properties of alkanes. One interesting problem on it is to find the maximum and minimum 2R value and its corresponding graphs among classes of graphs. In this talk, we will talk about the maximum and minimum 2R value on trees with fixed size.




Abstracts, meeting USC Sumter February 19 2010


Anthony Coyne

Mathematics, Metaphysics, Morality

Reflection on what Plato says about the place of the study of mathematics in the education of the guardians provides a sharp contrast with what is happening as USC approaches revising general education.  In particular, Plato does not advocate the study of mathematics because it leads to any of the learning outcomes identified by USC.  USC hopes students will be able to apply the methods of mathematics, statistics, or analytical reasoning to critically evaluate data, solve problems, and effectively communicate findings verbally and graphically. Plato hopes instead for  more important psychological, metaphysical, moral and political outcomes.

Wei-Tian Li:

Lubell Function and Forbidden Subposets

In 1928, Sperner proved that the size of a largest antichain in the Boolean lattice Bn is equal to n choose n/2. Since then, the results on largest sizes families not containing some specific posets were discovered gradually. The well-known inequality, the LYM-inequality, was individually used by Lubell, Yamamoto, and Meshalkin to reprove Sperner's Theorem. We will introduce the Lubell function, derived from LYM-inequality, and use it to estimate the maximal sizes of families do not contain some posets P. (This is a joint work with Jerrold R. Griggs and Lincoln Lu.)


Wei-Kai Lai:

Banach Space, Riesz Space, and Banach Lattice

In many Functional Analysis books, we can find the following definitions: a Banach space is a complete normed linear space; a Riesz space is an ordered vector space with the lattice structure. Rephrasing them with earthly language, you will be able to measure the distance of two objects in a Banach space and you will be able to compare which one is bigger among two objects in a Riesz space. In this talk, I will introduce some basic examples of these two spaces, and together with their properties. Finally, I will introduce a special space with both structures, called Banach Lattice.

Andrew Yingst:

Partition Polynomials And Their Uses (part II)

If a coin has probability x of coming up heads, and E is a set of possible outcomes of finitely many tosses of this coin, then the probability of event E is a polynomial in x, referred to as a partition polynomial.  We define and characterize these polynomials, and discuss some questions of dynamics on Cantor space that this characterization can be used to answer.

Charlie Cook

The “Magicness” of Powers of Some Magic Squares

(This paper is in collaboration with Michael R. Bacon and Rebecca A. Hillman)

Several Powers of a variety of additive magic squares are computed and conditions which guarantee that they are also magic are investigated.

Abstracts, meeting USC Lancaster October 30 2009



Andrew Yingst:

Partition Polynomials And Their Uses

If a coin has probability x of coming up heads, and E is a set of possible outcomes of finitely many tosses of this coin, then the probability of event E is a polynomial in x, referred to as a partition polynomial.  We define and characterize these polynomials, and discuss some questions of dynamics on Cantor space that this characterization can be used to answer.

Leandro Junes

From Hyperplanes to Oriented Matroid Programs.

:I will discuss the generalization of the simplex algorithm in Oriented Matroids. This will lead to an important class of Oriented Matroids called Euclidean Oriented Matroids.

Rigoberto Florez

Some topics in harmonic matroids

A matroid is a combinatorial generalization of the linear independence concept. There are no linear relations, only dependent and independent sets. Many geometric properties extend to matroids. The harmonic conjugation extends from projective geometry to matroids. In this talk I am going to discuss some results in harmonic matroids

Alexandru Gabriel Atim

On balanced property of words over finite alphabet (part II)

Sturmian words are infinite words over a binary alphabet that has exactly n+1 factors of length n for every n.  Morse-Heldun Theorem states that this equivalent to the set of all balanced aperiodic words. In this talk we will generalize the balanced property on words and then we will classify all words having this property.




Abstracts, meeting USC Sumter September 18 2009


Alexandru Gabriel Atim

On balanced property of words over finite alphabet

Sturmian words are infinite words over a binary alphabet that has exactly n+1 factors of length n for every n.  Morse-Heldun Theorem states that this equivalent to the set of all balanced aperiodic words. In this talk we will generalize the balanced property on words and then we will classify all words having this property.  

Rebecca Hillman

On Products of Fibonacci Numbers and Their Recurrence Relations

Various products of Fibonacci numbers and their generalizations are investigated and recurrence relations for these products are obtained

Jason Holt

Estimates for the negative eigenvalues for a Random Schroedinger Operator with a decaying perturbation

We will discuss a one dimensional random Schoedinger Operator with a decaying perturbation.  We will give sufficient condition for the existence of finitely many negative eigenvalues.  The model is closely related to the one dimensional motion of a quantum particle in a crystal lattice.

Leandro Junes

What is an oriented matroid?

An oriented matroid is a combinatorial structure inspired by a hyperplane arrangement. In this talk I will give an informal definition of  oriented matroid from a Geometric view point. I will also discuss its interrelations with Graph Theory, Linear Programming and Topology.



  The views and opinions expressed in this page are strictly those of Rigoberto Florez. The contents of this page have not been reviewed or approved by the University of South Carolina.