Search

International audience; In this paper, we describe the plans for the first LongEval CLEF 2023 shared task dedicated to evaluating the temporal persistence of Information Retrieval (IR) systems and Text Classifiers. The task is motivated by recent research showing that the performance of these models drops as the test data becomes more distant, with respect to time, from the training data. LongEval differs from traditional shared IR and classification tasks by giving special consideration to evaluating models aiming to mitigate performance drop over time. We envisage that this task will draw attention from the IR community and NLP researchers to the problem of temporal persistence of models, what enables or prevents it, potential solutions and their limitations.

We present a wavenumber-explicit convergence analysis of the hp finite element method applied to a class of heterogeneous Helmholtz problems with piecewise analytic coefficients at large wavenumber $k$. Our analysis covers the heterogeneous Helmholtz equation with Robin, exact Dirichlet-to-Neumann, and second order absorbing boundary conditions, as well as perfectly matched layers.

This work is dedicated to the analysis of generalized perfectly matched layers (PMLs) for 2D electromagnetic wave propagation in dispersive waveguides. Under quite general assumptions on frequency-dependent dielectric permittivity and magnetic permeability we prove convergence estimates in homogeneous waveguides and show that the PML error decreases exponentially with respect to the absorption parameter and the length of the absorbing layer. The optimality of this error estimate is studied both numerically and analytically. Finally, we demonstrate that in the case when the waveguide contains a heterogeneity supported away from the absorbing layer, instabilities may occur, even in the case of the non-dispersive media. Our findings are illustrated by numerical experiments.

Backward stochastic differential equations extend the martingale representation theorem to the nonlinear setting. This can be seen as path-dependent counterpart of the extension from the heat equation to fully nonlinear parabolic equations in the Markov setting. This paper extends such a nonlinear representation to the context where the random variable of interest is measurable with respect to the information at a finite stopping time. We provide a complete wellposedness theory which covers the semilinear case (backward SDE), the semilinear case with obstacle (reflected backward SDE), and the fully nonlinear case (second order backward SDE).

The theory of backward SDEs extends the predictable representation property of Brownian motion to the nonlinear framework, thus providing a path-dependent analog of fully nonlinear parabolic PDEs. In this paper, we consider backward SDEs, their reflected version, and their second-order extension, in the context where the final data and the generator satisfy L1-type of integrability condition. Our main objective is to provide the corresponding existence and uniqueness results for general Lipschitz generators. The uniqueness holds in the so-called Doob class of processes, simultaneously under an appropriate class of measures. We emphasize that the previous literature only deals with backward SDEs, and requires either that the generator is separable in (y,z), see Peng [Pen97], or strictly sublinear in the gradient variable z, see [BDHPS03], or that the final data satisfies an LlnL-integrability condition, see [HT18]. We by-pass these conditions by defining L1-integrability under the nonlinear expectation operator induced by the previously mentioned class of measures.

We consider a moral hazard problem with multiple principals in a continuous-time model. The agent can only work exclusively for one principal at a given time, so faces an optimal switching problem. Using a randomized formulation, we manage to represent the agent's value function and his optimal effort by an It\^o process. This representation further helps to solve the principals' problem in case we have infinite number of principals in the sense of mean field game. Finally the mean field formulation is justified by an argument of propagation of chaos.

An important problem from invariant theory is to describe the subspace of a tensor power of a representation invariant under the action of the group. According to Weyl's classic, the first main (later: 'fundamental') theorem of invariant theory states that all invariants are expressible in terms of a finite number among them, whereas a second main theorem determines the relations between those basic invariants.Here we present a transparent, combinatorial proof of a second fundamental theorem for the defining representation of the symplectic group $Sp(2n)$. Our formulation is completely explicit and provides a very precise link to $(n+1)$-noncrossing perfect matchings, going beyond a dimension count. As a corollary, we obtain an instance of the cyclic sieving phenomenon., Un problème important de la théorie des invariantes est de décrire le sous espace d’une puissance tensorielle d’une représentation invariant à l’action du groupe. Suivant la classique de Weyl, le théorème fondamental premier pour la représentation standard du groupe sympléctique dit que tous les invariants peuvent être exprimés entre un nombre fini d’entre eux. Par ailleurs, un théorème fondamental second détermine les relations entre ces invariants basiques.Ici, nous présentons une preuve transparente d’un théorème fondamental second pour la représentation standard du groupe sympléctique $Sp(2n)$. Notre formulation est complètement explicite et elle fournit un lien très précis avec les couplages parfaits $(n+1)$ -noncroissants, plus précis qu’un dénombrement de la dimension. Comme corollaire nous exhibons un phénomène de crible cyclique.

The descent set of an oscillating (or up-down) tableau is introduced. This descent set plays the same role in the representation theory of the symplectic groups as the descent set of a standard tableau plays in the representation theory of the general linear groups. In particular, we show that the descent set is preserved by Sundaram's correspondence. This gives a direct combinatorial interpretation of the branching rules for the defining representations of the symplectic groups; equivalently, for the Frobenius character of the action of a symmetric group on an isotypic subspace in a tensor power of the defining representation of a symplectic group., Dans cet article, nous définissons la notion d’ensemble de descentes pour un tableau oscillant. Ces descentes sont analogues aux descentes d’un tableau standard dans la théorie des représentations des groupes généraux linéaires. Nous montrons que la correspondance de Sundaram préserve cet ensemble et nous donnons une interprétation combinatoire directe des règles de branchement pour la représentation des groupes symplectiques. Enfin, nous décrivons combinatoirement les caractères de Frobenius associés à l’action du groupe symétrique sur les composantes isotypiques du produit tensoriel des représentations d’un groupe symplectique.