Past

The Jacobi-Davidson method, proposed by Sleijpen and Van der Vorst more than a decade ago, has been successfully used to numerically solve large matrix eigenvalue problems. In this talk we will give an introduction to and an overview of this method, and also point out some recent developments.

We consider the hedging of a claim on a non-traded asset using a correlated traded asset, when the agent does not know the true values of the asset drifts, a partial information scenario. The drifts are taken to be random variables with a Gaussian prior distribution. This is updated via a linear filter. The result is a full information model with random drifts. The utility infdifference price and hedge is characterised via the dual problem, for an exponential utility function. An approximation for the price and hedge is derived, valid for small positions in the claim. The effectiveness of this hedging strategy is examined via simulation experiments, and is shown to yield improved results over the Black-Scholes strategy which assumes perfect correlation.

We propose and analyze a fully discrete Fourier collocation scheme to

solve numerically a nonlinear equation in 2D space derived from a

pattern forming gradient flow. We prove existence and uniqueness of the

numerical solution and show that it converges to a solution of the

initial continuous problem. We also derive some error estimates and

perform numerical experiments to illustrate the theory.

In this talk we will begin by discussing the notion of homogenization as an extension to the continuum assumption and regimes in which it breaks down. We then discuss various approaches to dealing with randomness whilst determining the effective properties of acoustic, thermal and elastic media.  In particular we show how the effective properties depend on the randomness of the microstructure

``The Utility of Wealth,'', Markowitz's ``other'' 1952 paper, explains observed risk-seeking and risk-avoidance behaviour by a utility function which has deviation from customary wealth, rather than wealth itself, as its argument. It also assumes that utility is bounded above and below.

This talk presents a class (GUW) of functions which generalise

utility-of-wealth (UW) functions. Unlike the latter functions, the

class is too broad to have interesting, verifiable implications. Rather, various subclasses have such implications. A recent paper by Gillen and Markowitz presents notations to specify various subclasses, and explores the properties of some of these.

This talk extends this classification of risk-facing behaviour to non-utility-maximising behaviour as described by Allais and Ellsberg, and formalised by Mark Machina.

The classical expected utility maximisation theory for financial asset allocation is premised on the assumption that human beings when facing risk make rational choices. The theory has been challenged by many observed and repeatable empirical patterns as well as a number of famous paradoxes and puzzles. The prospect theory in behavioural finance use cognitive psychological techniques to incorporate anomalies in human judgement into economic decision making. This lecture explains the interplay between risk and human judgement, and its impact on dynamic asset allocation via mathematically establishing and analysing a behavioural portfolio choice model.

The geometric Langlands program aims at a "spectral decomposition" of certain derived categories, in analogy with the spectral decomposition of function spaces provided by the Fourier transform. I'll explain such a geometrically-defined spectral decomposition of categories for a particular geometry that arises naturally in connection with integrable systems (more precisely, the quantum Calogero-Moser system) and representation theory (of Cherednik algebras). The category in this case comes from the moduli space of vector bundles on a curve equipped with a choice of ``mirabolic'' structure at a point. The spectral decomposition in this setting may be understood as a case of ``tamely ramified geometric Langlands''. In the talk, I won't assume any prior familiarity with the geometric Langlands program, integrable systems or Cherednik algebras.

I will give a characterization of connected Lie groups admitting a quasi-isometric embedding into a CAT(0) metric space. The proof relies on the study of the geometry of their asymptotic cones

In the first part of the talk, we fit the Hua-Pickrell measure (which is a two parameters deformation of the Haar measure) on the unitary group and the Ewens measure on the symmetric group in a same framework. We shall see that in the unitary case, the eigenvalues follow a determinantal point process with explicit hypergeometric kernels. We also study asymptotics of these kernels. The techniques used rely upon splitting of the Haar measure and sampling techniques. In the second part of the talk, we provide a matrix model for the circular Jacobi ensemble, which is the sampling used for the Hua-Pickrell measure but this time on Dyson's circular ensembles. In this case, we use the theory of orthogonal polynomials on the unit circle. In particular we prove that when the parameter of the sampling grows with n, both the spectral measure and the empirical spectral measure converge weakly in probability to a non-trivial measure supported only by one piece of the unit circle.

We consider a critical, measure-valued branching diffusion ξ in Rd, where the branching is continuous and the spatial motion is given by the heat flow. For d ≥ 2 and fixed t > 0, ξt is known to be an a.s. singular random measure of Hasudorff dimension 2. We explain how it can be approximated by Lebesgue measure on ε-neighbourhoods of the support. Next we show how ξt can be approximated in total variation near n points, and how the associated Palm distributions arise in the limit from elementary conditioning. Finally we hope to explan the duality between moment and Palm measures, and to show how the latter can be described in terms of discrete “Palm trees.”

In recent years Schanuel’s Conjecture (SC) has played a fundamental role

in the Theory of Transcendental Numbers and in decidability issues.

Macintyre and Wilkie proved the decidability of the real exponential field,

modulo (SC), solving in this way a problem left open by A. Tarski.

Moreover, Macintyre proved that the exponential subring of R generated

by 1 is free on no generators. In this line of research we obtained that in

the exponential ring $(\mathbb{C}, ex)$, there are no further relations except $i^2 = −1$

and $e^{i\pi} = −1$ modulo SC. Assuming Schanuel’s Conjecture we proved that

the E-subring of $\mathbb{R}$ generated by $\pi$ is isomorphic to the free E-ring on $\pi$.

These results have consequences in decidability issues both on $(\mathbb{C}, ex)$ and

$(\mathbb{R}, ex)$. Moreover, we generalize the previous results obtaining, without

assuming Schanuel’s conjecture, that the E-subring generated by a real

number not definable in the real exponential field is freely generated. We

also obtain a similar result for the complex exponential field.

Many portfolio optimization problems are directly or indirectly concerned with the current maximum of the underlying. For example, loockback or Russian options, optimization with max-drawdown constraint , or indirectly American Put Options, optimization with floor constraints.

The Azema-Yor martingales or max-martingales, introduced in 1979 to solve the Skohorod embedding problem, appear to be remarkably efficient to provide simple solution to some of these problems, written on semi-martingale with continuous running supremum.

The phenomenon of ignition is one with which we are all familiar, but which is remarkably difficult to define and model effectively. My own (description rather than definition) is “initiation of a (high temperature) self-sustaining exothermic process”; it may of course be desirable, as in your car’s engine, or highly undesirable, as the cause of many disastrous fires and explosions Both laboratory experiments and numerical simulations demonstrate its extreme sensitivity to external influences, past history and process (essentially chemical) kinetics, but at the heart of all instances there appears to be some “critical” unstable equilibrium state. Though some analytical modelling has been useful in particular cases, this remains in general virgin territory for applied mathematicians – perhaps there is room for some “knowledge transfer” here.

We recall that an elliptic curve is a curve of genus one with a rational point on it. Certain algorithms for determining the structure of the group of rational points on an elliptic curve produce a whole set of curves of genus one and then require that we determine which of these curves has a rational point.

Unfortunately no algorithm which has been proved to terminate is known for doing this. Such an algorithm or proof would probably have profound implications for the study of elliptic curves and may shed light on the Birch and Swinnerton-Dyer conjecture.

This talk will be about joint work with Samir Siksek (Warwick) on the development of a new algorithmic criterion for determining that a given curve of genus one has no rational points. Both the theory behind the criterion and recent attempts to make the criterion computationally practical, will be detailed.

Hilbert schemes classify subschemes of a given projective variety / scheme. They are special cases of Quot schemes which are moduli spaces for quotients of a fixed coherent sheaf. Hilb and Quot are among the first examples of moduli spaces in algebraic geometry, and they are crucial for solving many other moduli problems. I will try to give you a flavour of the subject by sketching the construction of Hilb and Quot and by discussing the role they play in applications, in particular moduli spaces of stable curves and moduli spaces of stable sheaves.

> There is currently tremendous interest in geometric PDE, due in part
> to the geometric flow program used successfully to attack the Poincare
> and Geometrization Conjectures.  Geometric PDE also play a primary
> role in general relativity, where the (constrained) Einstein evolution
> equations describe the propagation of gravitational waves generated by
> collisions of massive objects such as black holes.
> The need to validate this geometric PDE model of gravity has led to
> the recent construction of (very expensive) gravitational wave
> detectors, such as the NSF-funded LIGO project.  In this lecture, we
> consider the non-dynamical subset of the Einstein equations called the
> Einstein constraints; this coupled nonlinear elliptic system must be
> solved numerically to produce initial data for gravitational wave
> simulations, and to enforce the constraints during dynamical
> simulations, as needed for LIGO and other gravitational wave modeling efforts.
>
> The Einstein constraint equations have been studied intensively for
> half a century; our focus in this lecture is on a thirty-year-old open
> question involving existence of solutions to the constraint equations
> on space-like hyper-surfaces with arbitrarily prescribed mean
> extrinsic curvature.  All known existence results have involved
> assuming either constant (CMC) or nearly-constant (near-CMC) mean
> extrinsic curvature.
> After giving a survey of known CMC and near-CMC results through 2007,
> we outline a new topological fixed-point framework that is
> fundamentally free of both CMC and near-CMC conditions, resting on the
> construction of "global barriers" for the Hamiltonian constraint.  We
> then present such a barrier construction for case of closed manifolds
> with positive Yamabe metrics, giving the first known existence results
> for arbitrarily prescribed mean extrinsic curvature.  Our results are
> developed in the setting of a ``weak'' background metric, which
> requires building up a set of preliminary results on general Sobolev
> classes and elliptic operators on manifold with weak metrics. 
> However, this allows us to recover the recent ``rough'' CMC existence
> results of Choquet-Bruhat
> (2004) and of Maxwell (2004-2006) as two distinct limiting cases of
> our non-CMC results.  Our non-CMC results also extend to other cases
> such as compact manifolds with boundary.
>
> Time permitting, we also outline some new abstract approximation
> theory results using the weak solution theory framework for the
> constraints; an application of which gives a convergence proof for
> adaptive finite element methods applied to the Hamiltonian constraint.

We compute the high frequency limit of the Hemholtz equation with source term, in the case of a refraction index that is discontinuous along a sharp interface between two unbounded media. The asymptotic propagation of energy is studied using Wigner measures. First, in the general case, assuming some geometrical hypotheses on the index and assuming that the interface does not capture energy asymptotically, we prove that the limiting Wigner measure satisfies a stationary transport equation with source term. This result encodes the refraction phenomenon. Second, we study the particular case when the index is constant in each media, for which the analysis goes further: we prove that the interface does not capture energy asymptotically in this case.

Euler equations for incompressible fludis describe the evolution of the divergence-free velocity of a non-viscous fluid (when viscosity is present, we have the well-known Navier-Stokes equations). V. Arnold discovered that they correspond to geodesic equations in the space of volume-preserving diffeomorphisms but several exemples show that it is not always possible to solve the corresponding variational problems inducing minimal energy displacements. A solvable relaxed version, in a non-deterministic setting (measures on the path space, with possible splitting of the particles), has been introduced by Y. Brenier who intensively studied the problem. Together with M. Bernot and A. Figalli we founded new solutions and characterization results. In the talk I'll present the most interesting features of the problem and of its solutions.

Mike Giles recently came up with a very general technique that improves the fundamental complexity of Monte Carlo simulation in the context where stochastic differential equations are simulated numerically. I will discuss some work with Mike Giles and Xuerong Mao that extends the theoretical support for this approach to the case of financial options without globally Lipschitz payoff functions. I will also suggest other application areas where this multi-level approach might prove valuable, including stochastic computation in cell biology.