Past

The aim of this talk is to render the power of (short-step) interior-point methods for linear programming (and by extension, convex programming) intuitively understandable to those who have a basic training in numerical methods for dynamical systems solving. The connection between the two areas is made by interpreting line-search methods in a forward Euler framework, and by analysing the algorithmic complexity in terms of the stiffness of the vector field of search directions. Our analysis cannot replicate the best complexity bounds, but due to its weak assumptions it also applies to inexactly computed search directions and has explanatory power for a wide class of algorithms.

Co-Author: Coralia Cartis, Edinburgh University School of Mathematics.

It is well-known that Euclidean domains are PIDs; examples proving that the inclusion is strict are not commonly known. Here is one.

The space of smooth rational curves of degree d in projective space admits various moduli theoretic compactifications via GIT, stable maps, stable sheaves, Hilbert scheme and so on. I will discuss how these compactifications are related by explicit blow-ups and -downs for d

Consider the Erdos-Renyi random graph $G(n,p)$ inside the critical window, so that $p = n^{-1} + \lambda n^{-4/3}$ for some real \lambda. In

this regime, the largest components are of size $n^{2/3}$ and have finite surpluses (where the surplus of a component is the number of edges more than a tree that it has). Using a bijective correspondence between graphs and certain "marked random walks", we are able to give a (surprisingly simple) metric space description of the scaling limit of the ordered sequence of components, where edges in the original graph are re-scaled by $n^{-1/3}$. A limit component, given its size and surplus, is obtained by taking a continuum random tree (which is not a Brownian continuum random tree, but one whose distribution has been exponentially tilted) and making certain natural vertex identifications, which correspond to the surplus edges. This gives a metric space in which distances are calculated using paths in the original tree and the "shortcuts" induced by the vertex identifications. The limit of the whole critical random graph is then a collection of such

metric spaces. The convergence holds in a sufficiently strong sense (an appropriate version of the Gromov-Hausdorff distance) that we are able to deduce the convergence in distribution of the diameter of $G(n,p)$, re-scaled by $n^{-1/3}$, to a non-degenerate random variable, for $p$ in the critical window.

This is joint work (in progress!) with Louigi Addario-Berry (Universite de Montreal) and Nicolas Broutin (INRIA Rocquencourt).

I will review our current mathematical understanding of waves on black hole backgrounds, starting with the classical boundedness theorem of Kay and Wald on Schwarzschild space-time and ending with recent boundedness and decay theorems on a wider class of black hole space-times.

The talk will survey some recent and not so recent work on the

Hausdorff and box dimension of self-affine sets and related

attractors and repellers that arise in certain dynamical systems.

This talk will introduce Dirichlet's Theorem on the approximation of real numbers via rational numbers. Once this has been established, a stronger version of the result will be proved, viz Hurwitz's Theorem.

Random polymers are used to model various physical ( Ising inter- faces, wetting, etc.) and biological ( DNA denaturation, etc.) phenomena They are modeled as a one dimensional random walk (Xn), with excursion length distribution

P(E1 = n) = (n)=nc, c > 1, and (n) a slowly varying function. The polymer gets a random reward, whenever it visits or crosses an interface. The random rewards are realised as a sequence of i.i.d. variables (Vn). Depending on the relation be- tween the mean value of the disorder Vn and the temperature, the polymer might prefer to stick on the interface (pinning) or undergo a long excursion away from it (depinning).

In this talk we will review some aspects of random polymer models. We will also discuss in more detail the pinning-depinning transition of the 'Pinning' model and also its relation to other directed polymer models

We study the class of Azema-Yor processes which are of the form F(M_t)-f(M_t)(X_t-M_t), where F'=f, X_t is a semimartingale with no positive jumps and M_t is its running maximum. We show that these processes arise as unique strong solutions to the Bachelier SDE which we also show is equivalent to the DrawDown SDE. The proofs are greatly simplified thanks to (algebraic) group property of the set of AY processes indexed by functions. We then restrict our attention to the case when X is a martingale. It turns out that the AY martingales are the only local martingales of the form H(X_t,M_t) for a Borel function H. Furthermore, they can also be characterised by their optimal

properties: all uniformly integrable martingales whose maximum dominates a given target are dominated by an AY martingale in the concave ordering of terminal values. We mention how these results find direct applications in portfolio optimisation/insurance theory.

Joint work with Laurent Cararro and Nicole El Karoui

Since summer 2007 financial markets moved in unprecedented ways. Volatility was extremely high. Correlations across the board increased dramatically. More importantly, also much deeper fundamental changes took place. In this talk we will concentrate on the following two aspects, namely, inter-bank unsecured lending at LIBOR and 40% recovery.

Before the crisis it was very realistic for the banks to consider that risk free rate of inter-bank lending, and hence also of funding, is equivalent to 3M LIBOR. This logic was extended to terms which are multiples of 3M via compounding and to arbitrary periods by interpolation and extrapolation. Driven by advances in financial mathematics arbitrage free term structure models have been developed for pricing of interest rate exotics, like LIBOR Market Model (or BGM). We explain how this methodology was challenged in the current market environment. We also point to mathematical questions that need to be addressed in order to incorporate in the pre-crisis pricing and risk management methodology the current market reality.

We also discuss historically validated and universally accepted pre-crisis assumption of 40% recovery. We expose its inconsistency with the prices observed now in the structured credit markets. We propose ways of addressing the problem and point to mathematical questions that need to be resolved.

We present a topos-theoretic interpretation of (a categorical generalization of) Fraïssé's construction in Model Theory, with applications to countably categorical theories. The proof of our main theorem represents an instance of exploiting the interplay of syntactic, semantic and geometric ideas in the foundations of Topos Theory.

Layered (or laminated) structures are increasingly used in modern industry (e.g., in microelectronics and aerospace engineering). Integrity of such structures is mainly determined by the quality of their interfaces: poor adhesion or delamination can lead to a catastrophic failure of the whole structure. Can nonlinear waves help us to detect such defects?

We study the dynamics of a nonlinear longitudinal bulk strain wave in a split, layered elastic bar, made of nonlinearly hyperelastic Murnaghan material. We consider a symmetric two-layered bar and assume that there is perfect interface for x 0, where the x-axis is directed along the bar. Using matched asymptotic multiple-scales expansions and the integrability theory of the KdV equation by the Inverse Scattering Transform, we examine scattering of solitary waves and show that the defect causes generation of more than one secondary solitary waves from a single incident soliton and, thus, can be used to detect the defect.

The theory is supported by experimental results. Experiments have been performed in the Ioffe Institute in St. Petersburg (Russia), using holographic interferometry and laser induced generation of an incident compression solitary wave in two- and three-layered polymethylmethacrylate (PMMA) bars, bonded using ethyl cyanoacrylate-based (CA) adhesive.

In this talk I will discuss how to construct generalized traces

and modified dimensions in certain categories of modules. As I will explain

there are several examples in representation theory where the usual trace

and dimension are zero, but these generalized traces and modified dimensions

are non-zero. Such examples include the representation theory of the Lie

algebra sl(2) over a field of positive characteristic and of Lie

superalgebras over the complex numbers. In these examples the modified

dimensions can be interpreted categorically and are closely related to some

basic notions involving the representation theory. This joint work with Jon

Kujawa and Bertrand Patureau.