Combinatorics Seminar
Everybody Welcome!
Tuesday's in SE 215
April 17 , 2007, 3:15 - 4:00 pm: Aaron Meyerowitz
Tiling
Abstract:
A polyomino is a generalization of the domino to a collection of squares
of equal size arranged with coincident sides. The figure shows 76 copies
of a certain polyomino tiling a 19x28 rectangle. For a given polyomino one
can ask: Are any rectangles tiled? If so, which ones or what is the
smallest one (The given rectangle is minimal)? Related questions exist for
other types of tiles and regions.
April 3 , 2007, 3:15 - 4:00 pm: Shanzhen Gao
Catalan-like Numbers
Abstract:
We will present Catalan-like Numbers, Hankel Matrices,
determinants, binomial formulae, sum coefficients, paths and trees.
February 20, 2007: Shaun Sullivan
Eulerian Coefficients and Dyck Paths without Long Sequences of Down Steps
January 31: Ron Mullin and Melissa Berg
A Problem about Pairwise Balanced Designs of Drake and Larson
Abstract.
October 18: Heinrich Niederhausen
From Germain Kreweras to Mireille Bousquet-Melou
We count the number of lattice paths in the quarter plane that take n unit steps in the North-East, South, or West direction.
May 3: Predrag Cudic and Dragan Radulovic
Combinatorics and Combinatorial Chemistry
We will present a short introduction on molecular recognition
of
biologically important mono- and disaccharides.
Carbohydrate recognition
is
particularly interesting because of its important role in biological
systems. Carbohydrate recognition is known to mediate cell-cell
recognition,
including infection of cells by pathogens, the distribution and
reactivity
of proteins within cells, and many aspects of the immune response.
However,
carbohydrate recognition by natural systems is still poorly understood.
On
the other hand, studies on synthetic carbohydrate receptors could make
important contributions to a better understanding of basic principles
of
this process.
Synthetic carbohydrate receptors could be used as drugs
(e.g.
anti-infective agents), to target cell types (acting as "synthetic antibodies"), to transport saccharides or related pharmaceuticals
across
cell membranes or in carbohydrate sensors. We have successfully
synthesized
macrocyclic receptor molecules capable of binding various mono- and
disaccharides in water. Using the combinatorial chemistry approach
these
synthetic receptor molecules can be further optimized for very specific
carbohydrate binding. Next, we will translate the chemical ideas into
more
mathematically tractable problem.
The last 20 minutes will be reserved
for
brainstorming and "calls for collaborations". We believe that this could
be a
great little projects for our students, and a nice introduction into the
bio-mathematical world.
April 26: Spyros Magliveras
From Payley and Hadamard to Higman and Sims
Serious students of mathematics and even seasoned researchers have often
studied in depth, but separately, mathematical objects such as Hadamard matrices, codes, finite geometries & designs, strongly regular graphs & association
schemes, finite simple groups and their classification. Eventually, one may see
connections among these objects and between classes of objects, and these illuminate each subarea and provide a new view point for examining all objects
simultaneously. From now on the researcher will have a totally new way of
looking at new problems.
In this talk we present a transversal walk from the Payley construction to
the Golay code, to the large Mathieu groups and related Witt systems, to the
construction of the Higman-Sims graph on 100 vertices, and to the proof of
existence of the "new" sporadic simple group of Higman and Sims.
Key words: Hadamard matrices, binary linear codes, Golay code, Mathieu
groups, Steiner systems, Witt systems, strongly regular graphs, the Higman-
Sims graph, the Higman-Sims sporadic simple group.
April 19: Katherine Humphreys
Counting Lattice Paths with Infinite Step Sets
using Sheffer Sequences
April 12: Wandi Wei
The Super Vertex-Gracefulness of Some Cartesian Product Graphs
March 22, 29, and April 5: Navin Singhi, School of Mathematics, Tata Institute of Fundamental Research, Mumbai
Studying t- designs and other families of subsets of a finite set algebraically
In these three lectures a general
t-k existence problem and some algebraic methods developed to study such problems will be
described. Special cases of the problem include well known unsolved problems like characterising parameters of t- designs or
degree sequences of k-uniform hypergraphs. Some other closely related unsolved problems are characterising f- vectors of pure
simplicial complex, parameters of Partial Steiner Systems, order of projective planes etc. Proofs of some results like that of
Graham Li -Li on a basis for space of null t- designs or the well known λ large theorem of Wilson and Graver and Jurkat on
the existence of t- designs will be sketched.
Recently developed concept of tags on subsets of finite sets, which provides a
unified approach to some of these classical problems and results will be described.
March 15 : Vladimir Bozovic
The Distribution of the Size of the Intersection of k-Tuples of Intervals
March 1 : Aaron Meyerowitz
Tiling Questions
Here is a 12 by 20 rectangle tiled by 60
T shaped tiles. How many such tilings are there
for this and other rectangles? What do the other lines and curves mean? MAPLE is used both for illustrations
and calculations with symbolic transfer matrices. The number of tilings for this particular rectangle can
be obtained by substituting x=2 into a certain polynomial of degree 15. There turn out to be 350,950,482
tilings.
February 22: Ron Mullin
Conway's Checker-Jumping Game
This presentation is in the area of recreational mathematics. It is an expository
talk about Conway's checker-jumping game, (which is actually a problem in counting), and is meant to
introduce it to those unfamiliar with it.
The material is accessible to with a background in college algebra, yet it involves some novel and clever ideas.
February 8: Ron Mullin
Coding for the Error-free Channel
An old result in information theory investigates efficient coding for error-free
channels via uniquely decipherable coding. A necessary and sufficient condition (the Kraft inequality)
for a special type of encoding known as prefix codes is easily established by combinatorial methods.
However to determine the relation of prefix codes to general uniquely decipherable codes seems to require
an analytic trick. In this talk we give a standard derivation of the Kraft inequality. We then use the
fact that a certain generating function is analytic in a neighborhood of the origin to relate prefix codes
to uniquely decipherable codes.