Department Colloquium

Department of Mathematical Sciences
Georgia Southern University
Statesboro, GA 30458-8093

Spring 2005 Lectures

Summary
 
Date
Speaker
Title
Friday, January 28 Gary Gruenhage, Auburn University Frechet-Urysohn for Finite Sets
Friday, February 4 Frederic Mynard, Georgia Southern Univ. Stability of local topological properties under product
Friday, February 11 (3:00 p.m. - 5:00 p.m.) 1. Hakki Turgay Kaptanoglu,  University of Virginia
2. Semra Ozturk Kaptanoglu, University of Virginia
1. Function theory in diagonal Besov spaces

2. Betti numbers of fixed point sets

Thursday, February 17 (CLEC lecture, 7:00 p.m. - 8:00 p.m., Math/Physics 3001) Ralph Kopperman, City College of CUNY Not all points are created equal.
Friday, February 18 Ralph Kopperman, City College of CUNY Partial metrics
Friday, February 25 Haomin Zhou,  Georgia Institute of Technology Variation and PDE techniques in wavelet inpainting
Friday, March 4 ------------- ------------
Friday, March 11 Peter Nyikos, Univ. of South Carolina Elbow room in Banach spaces
Friday, March 25 Renato D.C. Monteiro, School of ISyE, Georgia Institute of Technology A Generic Iterative Solver-Based Infeasible Primal-Dual Path-Following 
Algorithm for Convex Quadratic Programming
Friday, April 1 Wen-Ran Zhang, Dept of Computer Science, Gerogia Southern Univ. YinYang Bipolar Sets and Dynamic Modus Ponens for Open World Reasoning
Friday, April 8 Len Olsen, Dept of Literature and Philosophy, Georgia Southern Univ. A Formal Ontology for a Theory of Notation And Some Issues Concerning Identity
Friday, April 15 Francois Ziegler, Georgia Southern Univ. Cotangents of the Skies: an Introduction to Symplectic Geometry
Friday, April 22 Yingkang Hu, Georgia Southern Univ. The Return of FoThe Return of Fortran -- Fortran 90/95/2003
Thursday, April 28 (2:30 p.m.- 5:00 p.m.) 1. (2:30 - 3:30) Tom Kunkle, College of Charleston

(3:30 - 4:00 Coffee Break)

2. (4:00 - 5:00) Carl de Boor, Univ. of Wisconsin, Madison (Distinguished Lecture Series)

1. Multivariate Interpolants with Bounded Derivatives
 
 

2. Ideal Interpolation

Friday, April 29 Krystyna Kuperberg , Auburn University Wild arcs in dynamics
Thursday, May 5 (11:00 a.m.- 12: 00 noon), 
MP 3311
Jun Luo, Zhongshan Univ. (currently visiting Queens College, CUNY) On boundary structure of plane tiles
Tuesday, June 28 (3:00 - 4:00 p.m.) CNRS Marseille and Hebrew University of Jerusalem Some Aspects of Diffeologies 

Distinguished Lecture Series--- New!!

Fall 2004 Lectures








Details of Spring 2005 lectures

1.  Friday, January 28, 2005, 3:00 p.m. - 4:00 p.m.  Room: MP 3314  (Visiting Lecture).
 

Speaker:  Gary Gruenhage, Auburn University.

Title: Frechet-Urysohn for finite sets.

Abstract: A fundamental convergence property of topological spaces is the following, called the Frechet-Urysohn property: whenever
a point p is in the closure of a set A, there is a sequence of points of A converging to p.  We will discuss a natural generalization of this
property which in effect replaces the points of A by finite sets.  This property has important connections to the analysis of the Frechet-Urysohn property in products, to convergence in topological groups, and to certain topological games.
 
 

2.  Friday, February 4, 2005,  Room: MP 3314
 

Speaker:  Frederic Mynard, Georgia Southern Univ.

Title:  Stability of local topological properties under product.

Abstract: Local topological properties can be interpreted as properties of neighborhood filters. A general scheme to study the stability under
product of various properties of filters is presented, and applied to topological product theorems.
 
 

3.  Friday, February 11, 2005, 3:00 p.m. - 5:00 p.m.  Room: MP 3314. (Visiting Lectures).
 

1. Speaker:  Hakki Turgay Kaptanoglu, Univ. of Virginia.

Title: Function theory in diagonal Besov spaces.

Abstract:   Diagonal Besov spaces Bpq of analytic functions on the disc or the ball generalize the weighted Bergman spaces
Aqp,  q>-1, to all real values of the parameter q.  As usual, 0<p<=infinity.  The  Bpq  spaces include many other
function spaces as special cases.  We first discuss elementary properties of functions in  Bpq .  Then we give a complete
description of bounded Bergman projections from the Lebesgue classes onto  Bpq  and their right inverses.  If time remains,
we mention Carleson measures and a characterization of Toeplitz operators on  Bpq .  The results are natural extensions of
those known for weighted Bergman spaces.
 
 

2. Speaker:  Semra Ozturk Kaptanoglu,  Univ. of Virginia.

Title: Betti numbers of fixed point sets.

Abstract: Let G be a group and  X  be a  space on which G acts. For  a subgroup L in G, the subspace  XL  is the set of points in X  fixed by the action of L.   Capturing information on  XL  from the equivariant cohomology of X is a problem attacked by Borel-Atiyah-Quillen-Hsiang using localization. On the other hand the G-action on X induces a kG-module structure on its cohomology groups in each dimension where K is a field and kG is the group algebra. For suitable G and X , we will show how to get information on the Betti numbers of  XL  using the kL-module structure of  M.
 
 

4.  CLEC Lecture. Thursday, February 17, 2005, 7:00 p.m. - 8:00 p.m.  Room: 3001 Math/Physics Building.

Speaker:  Ralph Kopperman, City College of CUNY.  (CLEC speaker).

Title: Not all points are created equal.

Abstract: The computer screen looks like a rectangle in the plane, but under a magnifying glass it breaks up into an array of about a million
bright dots (pixels). It could save memory space to store boundaries and colors of regions, rather than the color of each pixel. This turns out
to be impossible to do if the screen is thought of as a set of equal points, but possible and useful if it is thought of as a space where some
points are more equal than others.
 
 

5.  Friday, February 18, 2005, 3:00 p.m. - 4:00 p.m.  Room: MP 3314.

Speaker:  Ralph Kopperman, City College of CUNY

Title: Partial metrics.

Abstract: Generally accepted metric axioms, such as d(x,y)=d(y,x) and d(x,x)=0 seem "right", but there are
extremely natural situations in which they fail. We discuss these failures, the theory that arises from them,
and applications (in computing) of generalized metrics for which these axioms fail.
 
 

6.  Friday, February 25, 2005, 3:00 p.m. - 4:00 p.m.  Room: MP 3314 (Visiting Lecture).

Speaker:  Haomin Zhou, School of Mathematics, Georgia Tech.

Title: Variation and PDE Techniques in Wavelet Inpainting.

Abstract: In this talk, I will present a recent work (collaborated with Tony Chan (UCLA) and Jackie Shen (Minnesota)) on image inpainting in wavelet domain.  The problem is closely related to the classical image inpainting, with the difference being that the inpainting regions that we are interested in are in the wavelet domain, that brings new challenges to the reconstructions,  as there is no geometrically well defined inpainting region in the pixel  domain, and the damage is inhomogeneous. We propose new variational models, especially total variation minimization in conjunction with wavelets for the image inpainting problems in the wavelet domain.  The models lead to PDE's, which are Euler-Lagrange
equations of the variational formulations, in the wavelet domain and can be solved numerically. The proposed models can have effective and automatic control over geometric features of the inpainted images including sharp edges, even in the presence of substantial
loss of wavelet coefficients, including in the low frequencies.
 
 

7.  Friday, March 4 , 2005,  Room: MP 3314  (Cancelled due to job interview.)
 

8.  Friday, March 11, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314  (Visiting Lecture)
 

Speaker:  Peter Nyikos, Department of Mathematics, University of South Carolina

Title: Elbow room in Banach spaces

Abstract: Separable Banach spaces have the following unexpected property. If a family of subsets is locally finite in the norm then
each member can be expanded to a set that is open in the weak topology, in such a way that the family of expansions is still locally
finite in the norm. This seems to be new even for Hilbert space, where we can even go further and expand to a locally finite collection of sets open in the natural product topology (thinking of members of Hilbert space as functions from the natural numbers to the reals).  We can do this for all  lspaces except loo , which is not a separable Banach space.  In fact, there is a countable closed discrete subset of loo which cannot be expanded to a locally finite collection of weakly open sets. Other examples, counterexamples and an application to Erdos space will be presented.
 
 

9.  Friday, March 25, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314  (Visiting Lecture)
 

Speaker:  Renato D.C. Monteiro,  School of ISyE, Georgia Institute of Technology Tech

Title:  A Generic Iterative Solver-Based Infeasible Primal-Dual Path-Following Algorithm for Convex Quadratic Programming

Abstract:  We develop an interior-point primal-dual long-step path-following algorithm for convex quadratic programming (CQP) whose search directions are computed by means of an iterative linear system solver. We propose a new linear system, which we refer to as the augmented normal equation (ANE), to determine the primal-dual search directions. Since the condition number of the matrix associated with the ANE may become large for degenerate CQP problems, we provide a family of preconditioners (e.g. partial-update, maximum weight basis, etc.) to better condition this matrix. We establish a uniform bound on the number of iterations required for the iterative solver to obtain a sufficiently accurate solution to the ANE. Since the iterative solver can only generate an approximate solution to the ANE, this solution does not yield a primal-dual search direction satisfying all equations of the primal-dual Newton system. We propose a unified approach to compute an inexact primal-dual search direction so that the Newton equation corresponding to the primal residual is satisfied exactly, while the one corresponding to the dual residual contains a manageable error which allows us to establish a polynomial bound on the number of outer iterations of our method.
 
 

10.  Friday, April 1, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314
 

Speaker:  Wen-Ran Zhang, Dept of Computer Science, Gerogia Southern Univ.

Title: YinYang Bipolar Sets and Dynamic Modus Ponens for Open World Reasoning

Abstract: Yinyang bipolar sets, bipolar lattices, dynamic modus ponens, and equilibrium relations are introduced which form a theory of bipolar machinery. YinYang bipolar sets and modus ponens build a bridge from a linear, static, and closed world to a non-linear, dynamic, and open world of equilibria or non-equilibria, that provide an effective means for bipolar information fusion and visualization. It is also shown that equilibrium relations as bipolar generalization of equivalence relations induce hard partitions or bipolar sets. Application examples are illustrated in cognitive mapping and bipolar diagnostic analysis.
 
 

11.  Friday, April 8, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314
 

Speaker:  Len Olsen, Dept of Literature and Philosophy, Georgia Southern Univ.

Title: A Formal Ontology for a Theory of Notation And Some Issues Concerning Identity

Abstract: The central issue of this presentation concerns the identity conditions for entities.  The theory of notation, like all other theories, presupposes the existence of certain entities.  A catalogue of these entities (or types of entities) is called an ontology.  My talk will focus on the possibility of using mereology as the basis of a formal ontology for the theory of notation.  Mereology, otherwise known as the calculus of
individuals or the logic of parts and wholes, is offered as an alternative to set theory.
 
 
 

12.  Friday, April 15, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314
 

Speaker:  Francois Ziegler, Department of Mathematical Sciences, Georgia Southern Univ.

Title: Cotangents of the Skies: an Introduction to Symplectic Geometry.

Abstract: Symplectic geometry was born in the 19th century as the geometry of classical mechanics, ray optics, and line complexes. Much later, in the
1960s, mathematicians began to realize that it is also the geometry of quantum mechanics and representation theory. In this talk I will attempt
an elementary tour of these ideas and techniques, using as a running example the manifold of oriented straight lines in R3 (a.k.a. light rays).
 

13.  Friday, April 22, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314
 

Speaker:  Yingkang Hu, Department of Mathematical Sciences, Georgia Southern Univ.

Title: The Return of Fortran -- Fortran 90/95/2003.

Abstract: Fortran, which first appeared in the mid 1950's, is one of the oldest high-level languages, and is a very successful one. Some say its structure is superior for vector computation. But it took a long break after its ANSI 1978 version (best known as Fortran 77), while new languages, such as Java, had emerged, and other old ones, such as C/C++, had evolved greatly. Many think Fortran is a dinosaur.

Well, old things die hard. Fortran came back in three new versions, Fortran 90/95, and just last November, Fortran 2003, with lots of new features. It is now the best language for scientific computation in my opinion. (Java and C/C++ are better for system and Internet programming.) Some of the new features are:

1. Contains Fortran 77;
2. Allows free format. Counting columns is no longer needed;
3. Dynamic memory allocation. (Arrays can be allocated after the program inputs their dimensions);
4. Recursive calling of subprograms;
5. High-level operations, such as matrix assignment/addition/multiplication, dot product, norm, and max/min of vectors/matrices, are intrinsic now;
6. It is modular;
7. Pointers installed;
8. Object-Oriented Programming is supported by Fortran 2003.

Code for matrix/vector calculation can be shortened greatly, but still remains readable. This is one of the main reasons people love Matlab, it's no longer a monopoly.

14. Thursday, April 28, 2005,   Room: MP 3314

1. Time: 2:30 p.m. - 3:30 p.m. (Visiting Lecture)

Speaker:  Tom Kunkle, College of Charleston

Title: Multivariate Interpolants with Bounded Derivatives.

Abstract: Let n and d be natural numbers and consider the following problem.  Construct a smooth  interpolant  F to function values f given at points m in Rd, where F depends locally and linearly on f, and F's derivatives of total degree n are bounded by a constant C times the corresponding divided differences of f.  Here C may depend on d and n, but must be independent of f and m.

Favard [1] gave an optimal solution to the  problem in case d=1 by bounding each of two consecutive derivatives by the corresponding divided differences and allowing m to be any discrete set in R. Such freedom would not be possible in the multivariate setting. In fact, if m is a tensor product grid in Rd, i.e., the Cartesian product of d increasing sequences of real numbers,  then such an interpolation scheme exists if and only if m has uniform spacing in each of the coordinate directions.

When m is a subset of a tensor product grid with uniform spacing, the interpolation problem may or may not have a  solution, depending on the geometry of the subset.

[1] Favard, J., Sur l'interpolation, J. Math. Pures Appl. \/ 19 (1940), 281--300.
 
 

2. Time: 4:00 p.m. - 5:00 p.m.  Distinguished Lecture Series

Speaker:  Carl de Boor, Univ. of Wisconsin, Madison

Title: Ideal Interpolation

Abstract: Starting from G. Birkhoff's definition of `ideal interpolation' (as a linear projector whose kernel is a polynomial ideal), multivariate polynomial
interpolation is explored.
 

15.  Friday, April 29, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314 (Visiting Lecture)

Speaker:  Krystyna Kuperberg, Auburn University

Title: Wild arcs in dynamics

Abstract: One of the most interesting aspects of dynamical systems are the topological invariants, the geometric side of the configuration of trajectories, and the nature of the individual trajectories.

A trajectory of a flow on a 3-manifold is wild  if the closure of at least one of its two  semi-trajectories is a wild arc. A  trajectory is 2-wild if the closure of each its two semi-trajectories is a wild arc. For example, the closure of a 2-wild trajectory may be the wild arc described by Artin and Fox (see MR0027512). A 2-wild trajectory may be also homoclinic, with the limit sets equal and consisting of a fixed point.

We describe a method of embedding wild trajectories in flows on 3-manifolds and obtain the following:  every boundaryless 3-manifold admits a flow with a discrete set of fixed points and such that every non-trivial trajectory is 2-wild. Every closed 3-manifold admits a flow with exactly one fixed point and such that every non-trivial trajectory is homoclinic and 2-wild.
 

16.  Thursday, May 4, 2005,  11:00 a.m.- 12: 00 noon,  Room: MP 3311 (Visiting Lecture)
 

Speaker: Jun Luo, Zhongshan University, China (currently visiting Queens College, City Univ. of New York)

Title: On boundary structure of plane tiles

Abstract:  We consider the boundary of a planar tile or components of its interior. Particularly, we obtain local connectivity of the whole boundary
whenever the tile itself is connected and locally connected, as an application of a new characterization of local connectivity. Also, we will
comment on a necessary and sufficient condition for  interior components of a tile to be bounded by a simple closed curve.
 

17.  Tuesday, June 28, 2005,  3:00 p.m. - 4:00 p.m.,  Room: MP 3314

Speaker: Patrick Iglesias, CNRS Marseille and Hebrew University of Jerusalem

Title:  Some Aspects of Diffeologies

Abstract:  Diffeology is a differential geometric framework introduced to deal with differential objects for which topology is not of a big help.  It is generally presented as an extension of the classical theory of manifolds, but it is much more than that.  It is a completely new point of view on mathematical objects which support "differentiable parametrizations."

The diffeology category is extremely stable under standard operations of sums, products, restrictions and quotients, and its applications run from singular spaces to infinite dimensional spaces.

I will give a general presentation of the theory and try to illustrate it through the famous examples of "irrational tori."  An irrational torus is defined as the quotient of an ordinary torus by a proper dense subgroup.  Irrational tori are topologically trivial, but surprisingly rich from the diffeological viewpoint.
 
 

If you have any question regarding the colloquium, please e-mail Dr. Sze-Man Ngai: ngai@gsu.mat.GeorgiaSouthern.edu.
 

( Last updated: June 23,  2005.)