Pure Mathematics Seminars



16th January 2007
No Seminar

23rd January 2007
Gerald Williams (Kent)
"Free subgroups in generalizations of Coxeter groups"
Abstract:
We consider groups defined by presentations that generalize those of Coxeter groups, in particular the so-called Pride groups,
and the `groups defined by periodic paired relations'. We show that under a suitable non-positive curvature condition the groups
contain a non-abelian free subgroup.

30th January 2007
Andreas Enge (Ecole Polytechnique)
"An L(1/3) algorithm for the discrete logarithm problem in low degree curves"
(joint work with Pierrick Gaudry)
Abstract:
Starting with the work by Adleman, DeMarrais and Huang more than ten years ago, it has been well-established that the discrete logarithm problem in Jacobians of curves of high genus over finite fields is easier to solve than in elliptic curves of the same size. Letting
$L (\alpha, c) = e^{(c + o (1)) (g \log q)^\alpha (\log (g \log q))^{1 -\alpha}}$
denote the subexponential function with respect to the curve genus $g$ and the cardinality of the finite field $q$, the algorithms for computing discrete logarithms in high genus curves have a complexity of $L (1/2, c)$. This is in contrast on one hand to elliptic curves, for which so far only exponential algorithms are known except for some rare special cases, but also to the computation of discrete logarithms in finite prime fields, for which the number field sieve yields a more efficient algorithm in $L (1/3, c)$. We present the first subexponential algorithm with an exponent $\alpha < 1/2$ to attack the discrete logarithm problem in a class of algebraic curves. The curves are characterised by their total degree being relatively low with respect to their genus. The algorithm has a complexity of $L (1/3, c)$ for computing the group structure and discrete logarithms in the Jacobian.

6th February 2007
Karin Erdmann (Oxford)
"Schensted correspondence and Littelmann paths"
Abstract:
This is joint work with J.A. Green and M. Schocker. We study the Littelmann path model for the case gl_n. In this case, the paths become words, and we work with the combinatorics of words.

13th February 2007
Gavin Brown (Kent)
"Key varieties and surgery in birational geometry"
Abstract:
I will sketch the ingredients that go into the classification of complex algebraic varieties (that is, the solutions of polynomial equations over the complex numbers), concentrating especially on the methods available to transform one variety into another. While these blowups, flips and flops arise (in Mori theory, at least) from rather abstract considerations, I am interested in writing them concretely, using simple equations and group actions. The punchline is that there are some classes of straightforward-but-large varieties, the _key_ varieties, that house whole families of flips, etc. (in much the same way as the "cooling tower" quadric surface contains the various conic curves as sections).

20th February 2007
Benjamin Klopsch (Royal Holloway)
“Anosov diffeomorphisms and strongly hyperbolic elements in arithmetic subgroups of SL_n(R)”
Abstract:
I will talk about some ongoing work, motivated by a long standing problem in the theory of dynamical systems. In particular, I will explain how p-adic methods lead to the construction of elements in SL_n(Z) whose eigenvalues e_1, …, e_n generate a free abelian subgroup of rank n-1 in the multiplicative group of positive real numbers. This is a special instance of a more general theorem, asserting the existence of strongly hyperbolic elements in arithmetic subgroups of SL_n(R).

27th February 2007
Beth Holmes(Cambridge)
"Which elements are in the same subgroup?"
Abstract:
If we are given a set of elements from a group then we can ask whether any of them lie in the same proper subgroup. Considering this leads to several related questions. We will look at three of these, namely subgroup coverings, pairwise generating sets and spread.

6th March 2007
Alan Camina (UEA)
"Automorphisms of finite linear spaces"
Abstract:
We look at recent work on classifying those quasisimple groups of large rank which can act line-transitively. The emphasis will be on the techniques which enable us to use the knowledge of the structure of the groups rather than on technical things about the classical linear groups.
All terms WILL be explained and they are not difficult.

13th March 2007
Oliver Baues (TU Karlsruhe)
"Constructions of aspherical manifolds"
Abstract:
A manifold is called aspherical if its universal covering space is contractible. This is the case, for example, if the universal covering is homeomorphic to an Euclidean space.
Given an abstract group $\Gamma$, there is the basic question if it is possible to construct compact aspherical smooth manifolds with fundamental group $\Gamma$, and also to understand the geometric properties of such manifolds. Ideally, one would like to classify them up to homeomorphism or up to diffeomorphism.
For example, 'most' polycyclic groups $\Gamma$ appear as fundamental groups of so called solvmanifolds. Another type of examples which appear in geometry are the fundamental groups of locally symmetric spaces. We would like to discuss a method which allows to build 'mixed' examples from these basic building blocks. This construction corresponds to the notion of group extension on the level of the fundamental group, and it has many interesting geometric properties.

20th March 2007
Victor Flynn (New College, Oxford)
"Cycles of Covers"
Abstract:
When a direct attack fails to determine all rational points on a curve C, one can try to find an associated collection of curves, whose rational points cover those on C, and attempt to resolve these curves. I shall describe a special case of this technique, when it is possible to apply the process repeatedly. I shall give an example to show that some curves can even be resistant to arbitrarily many repeated applications.

27th March 2007
Tuvi Etzion (Technion, Haifa)
"Correction of Multi-dimensional Cluster Errors"
Abstract:
In current memory devices for advanced storage systems the information is stored in two or more dimensions. In such systems errors usually take the form of multi-dimensional bursts. Usually, a cluster of errors either will be affected by the position in which the error event occurred or will be of an arbitrary shape. In this talk we will give a brief introduction to error-correcting codes and burst-correcting codes and distinguish between one-dimensional codes and multi-dimensional codes. We improve the known lower bound (Reiger bound) on the redundancy of two-dimensional codes. We will consider another lower bound implied by the number of possible burst patterns. We show variety of constructions of two-dimensional codes with redundancy close to the the lower bound. In particular we give constructions, where the shape of the burst is rectangle, Lee sphere, or an arbitrary shape. Most of these constructions can be easily generalized for multi-dimensional codes. The techniques we are using are mostly a combination of algebra and combinatorics.


ALL WELCOME !


This page lists this term's Pure Mathematics Seminars. All are welcome. All seminars will take place in the McCrea Building, Room 219, at 4.00pm, unless stated otherwise. Tea is served before the seminar at 3.30 in Room 237 of the McCrea Building.

Seminars of the past years.

Useful links:
Info for the speaker.
National Rail enquiries.
An interactive local street map. (In the right corner is Egham station, in the left corner is Royal Holloway. You can zoom the map, or go the left/right etc..., and print.) Note that the official college recommendation is to go via Station Road to the roundabout and walk up Egham Hill (A30). Even though this appears to be a detour it may be the fastest way. If you try, on your risk, the direct route make sure you reach Ripley avenue and enter the College at the end of Lynwood avenue.
For other directions (car, bus, taxi or from Heathrow see the following directions.)

Campus Map (The maths department is in no 17)
For other maps.

For any further information, please contact Dr Yiftach Barnea,
Tel: 01784 414689,
email: y.barnea@rhul.ac.uk