Alessio Savini (Université de Genève)
Explicit cocycles on Furstenberg boundaries abstract
Abstract:
Given a center-free semisimple Lie group without compact factors, it is known that its continuous bounded cohomology can be computed by the complex of essentially bounded functions on the Furstenberg boundary. In order to study the nature of the comparison map, more recently Monod has proved that the complex of (unbounded) measurable functions on the same boundary computes the continuous cohomology, except possibly in some low degrees determined by the rank. We investigate this phenomenon for SL(3,K), with K real of complex, and for products of isometries of hyperbolic spaces and we exhibit an explicit family of cocycles in low degrees. As a consequence we show that the comparison map is an isomorphism in degree 3, which is known for SL(3,K) but new for the product case.This is a joint work with Michelle Bucher.
10:30 • Université de Genève, Conseil Général 7-9, Room 1-05
Cécile Gachet (Université Côte d\'Azur)
The cone conjecture for certain Calabi-Yau manifolds obtained as fiber products abstract
Abstract:
The cone conjecture is a long-lasting conjecture in birational geometry, and can be viewed as a conjectural counterpart to the cone theorem, for K-trivial varieties. It predicts for any K-trivial variety X the existence of a rational polyhedral fundamental domain for the action of the group Aut(X) on the cone Nefe(X). If the conjecture is true for a certain X, it has important consequences, such as the finiteness of real structures on X, or the fact that the group Aut(X)/Aut0(X) is finitely presented. In this talk, I present a construction inspired by the work of Schoen, Namikawa and Grassi-Morrison, of Calabi-Yau manifolds obtained as fiber products over the projective line, and prove the cone conjecture for these Calabi-Yau manifolds. This is joint work with Hsueh-Yung Lin and Long Wang.
10:30 • Universität Basel, Spiegelgasse 1, Seminarraum 00.003
Urban Ulrych (EPF Lausanne)
Global Currency Hedging with Ambiguity abstract
Abstract:
This paper addresses the problem of optimal currency allocation for a risk-and-ambiguity-averse international investor. A robust mean-variance model with smooth ambiguity preferences is used to derive the optimal currency exposure in closed form. In the theoretical part of the paper, we characterize the sample-efficient currency hedging demand as the solution to a generalized ridge regression. Through the lens of these results, we show that the investor\'s dislike for model uncertainty induces stronger currency hedging demand. The empirical analysis demonstrates how ambiguity leads to a larger estimation bias and simultaneously narrows the confidence interval of the sample efficient optimal currency exposure. The out-of-sample backtest illustrates that accounting for ambiguity enhances the stability of optimal currency allocation over time and significantly reduces portfolio volatility net of transaction costs.
12:00 • EPF Lausanne, UniL campus, Extranef 126
Yu Li (MPIM Bonn)
Integrable systems on the dual of nilpotent Lie subalgebras and T-Poisson cluster structure abstract
Abstract:
Let g be a semisimple Lie algebra and g = n \\oplus h \\oplus n_- a triangular decomposition. Motivated by a construction of Kostant-Lipsman-Wolf, we construct an integrable system on the dual space of n_- equipped with the Kirillov-Kostant Poisson structure. The Bott-Samelson coordinates on the open Bruhat cell (equipped with the standard Poisson structure) makes it into a symmetric Poisson CGL extension, hence giving rise to a T-Poisson seed on it. We explain a relation between our integrable system and this T-Poisson seed. This is joint work in progress with Yanpeng Li and Jiang-Hua Lu.
14:00 • Université de Genève, Conseil Général 7-9, Room 6-13
Prof. Dr. Georgios Moschidis (EPFL )
Weak turbulence for the Einstein--scalar field system abstract
Abstract:
In the presence of confinement, the Einstein field equations are expected to exhibit turbulent dynamics. One way to introduce confinement to the equations is by imposing a negative value for the cosmological constant. In this setting, the AdS instability conjecture claims the existence of arbitrarily small perturbations to the initial data of Anti-de Sitter spacetime which, under evolution by the vacuum Einstein equations with reflecting boundary conditions at conformal infinity, lead to the formation of black holes after sufficiently long time.In this talk, I will present a rigorous proof of the AdS instability conjecture in the setting of the spherically symmetric Einstein-scalar field system. The construction of the unstable initial data will require carefully designing a family of initial configurations of localized matter beams and estimating the exchange of energy taking place between interacting beams over long periods of time, as well as estimating the decoherence rate of those beams. I will also briefly discuss some other problems related to turbulent phenomena in the asymptotically AdS setting.
15:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Gilles Vilmart (Université de Genève)
Explicit stabilized integrators for stiff systems: from deterministic to stochastic, optimal control, or convex optimization problems abstract
Abstract:
Explicit stabilized integrators are an efficient and popular alternative to implicit or semi-implicit methods to avoid the severe timestep restriction faced by standard explicit integrators applied to stiff dissipative systems of differential equations.
In this talk, we explain how the versatility of explicit stabilized methods permits to consider not only dissipative systems of (deterministic or stochastic) differential equations, but also stiff optimal control problems and convex optimization problems.
This talk is based on joint works with Assyr Abdulle (EPFL), Ibrahim Almuslimani (Rennes), Armin Eftekhari (Umea), Bart Vandereycken (Geneva), and Konstantinos C. Zygalakis (Edinburgh).
16:15 • EPF Lausanne, Salle GA 3 21
Prof. Dr. Michael Bronstein (University of Oxford)
Abstract:
Abstract: The message-passing paradigm has been the “battle horse” of deep learning on graphs for several years, making graph neural networks a big success in a wide range of applications, from particle physics to protein design. From a theoretical viewpoint, it established the link to the Weisfeiler-Lehman hierarchy, allowing to analyse the expressive power of GNNs. We argue that the very “node-and-edge”-centric mindset of current graph deep learning schemes may hinder future progress in the field. As an alternative, we propose physics-inspired “continuous” learning models that open up a new trove of tools from the fields of differential geometry, algebraic topology, and differential equations so far largely unexplored in graph ML.
16:30 • ETH Zentrum, Building KO2, Room F 150
Dr. Bernadette J. Stolz (University of Oxford)
Applications of global and local persistent homology for the shape of biological data abstract
Abstract:
Topological data analysis (TDA) is an emerging mathematical field that uses topological and geometric approaches to quantify the “shape” of data. In the first part of this talk, I will showcase how persistent homology, a method from TDA, can be used to spatially characterise structural abnormality in tumour blood vessel networks reconstructed from experimental data. More specifically, I will show that the number of vessel loops and their spatial distribution in these networks change over time when tumours undergo treatment with vascular targeting agents and radiation therapy. I will also highlight what insight TDA can give when applied to synthetic data generated from mathematical models of tumour-induced vascular growth. In the second part of the talk, I will demonstrate applications of local persistent homology. I will show how local persistent homology can be used to select landmarks from large and noisy data sets. In contrast to existing methods, this subsampling process is robust to outliers and is developed specifically as a preprocessing step for persistent homology. Based on similar ideas, I will introduce a novel method that can detect geometric anomalies, such as intersections or boundaries, in point cloud data sampled from intersecting surfaces. This detection is based on the computation of persistent homology in local annular neighbourhoods around points and is less sensitive to the size of the local neighbourhood and surface curvature than local principal component analysis.
17:15 • Université de Fribourg, room Phys 2.52