Abstract:
This workshop brings together dynamicists who use techniques from, and are motivated by, combinatorics, geometry, and number theory. The topics cover a broad swath of dynamics, including ergodic theory, homogeneous dynamics, symbolic dynamics, Teichmuller theory, and billiards, and the aim of the workshop is to highlight the common features across these subjects.
08:00 • EPF Lausanne, Room GA 3 21
Craig Evans (University of California, Berkeley)
Weak Convergence Methods for Nonlinear Partial Differential Equations abstract
Abstract:
This course will be an ambitious survey of rigorous methods for understanding solutions (u^\\epsilon) of various nonlinear PDE in various asymptotic limits. Assuming in particular that the (u^\\epsilon) converge weakly as (\\epsilon \\to 0), we want to identify what PDE the weak limit (u) solves. This can be a hugely complicated problem, since the weak convergence is usually incompatible with the nonlinearities; but a rich variety modern techniques can handle these issues for many interesting cases. The lectures will discuss recent developments concerning
(i) asymptotics for ODE,
(ii) maximum principle methods,
(iii) convexity and monotonicity,
(iv) oscillations and compensated compactness, and
(v) defect measures, with many examples and applications.
10:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Bernard Bonnard (Université de Dijon)
Contrôle en Temps Minimal des Réseaux Chimiques en Variant la Température et Application au Schéma de McKeithan abstract
Abstract:
Dans cet exposé on présente le problème de maximiser la production d’un réacteur chimique de type batch en controlant la température (Problème de Mayer). Le système se modélise en utilisant la théorie développée dans les années 70 par Feinberg, Horn, Jackson, à partir du graphe associé aux réactions. Le problème de Mayer peut être formulé comme un problème de temps minimal dont la cible est une variété terminale de codimension un. En utilisant le principe du Maximum de Pontriaguine les candidats optimaux sont à déduire de l’analyse des courbes extrémales définies par une dynamique Hamiltonienne. En utilisant la théorie géométrique du contrôle et la théorie des singularités, les synthèses temps minimales (feedback optimal)sont classifiées au voisinage de la variété terminale sous des conditions génériques dépendant des crochets de Lie du système. Pour cela on construit des formes semi-normales algébriques qui permettent d’évaluer les strates de la surface de commutation. La synthèse globale peut alors être calculée numériquement en propageant les singularités le long de la dynamique Hamiltonienne et en calculant les points focaux. Cette théorie est testée pour le schéma de McKeihtan dans le cas de deux réactions. Cela montre la difficulté calculatoire du problème même réduit au voisinage de la variété terminale. L’autre problème crucial étant d’avoir une borne globale sur le nombre de commutations des trajectoires optimales.
14:00 • Université de Genève, Section Mathématiques, Centre Acacias II, Salle 623
Reda Chhaibi (Université de Toulouse)
Quantum SL_2, infinite curvature and Pitman\'s 2M-X Theorem abstract
Abstract:
Pitman\'s theorem (1975) is an aesthetic theorem from probability theory, with geometry and representation theory related to SL_2, in disguise. Many proofs do exist, and the goal of this talk is to present a unified point of view regarding two proofs.
- A proof by Bougerol and Jeulin - which generalizes to all semi-simple groups. They consider a Brownian motion on the symmetric space $H^3 = SL_2(\\mathbb{C})/SU_2$, with varying curvature r and then take the limit $r \\rightarrow \\infty$.
- Biane defined and studied quantum random walks on the enveloping algebra of SL_2. Then, he made the connection to the representation theory of the Jimbo-Drinfeld quantum group $\\mathcal{U}_q(sl_2)$, in the crystal regime, i.e $q \\rightarrow 0$.
Why should the crystal regime $q=0$ for quantum groups be related infinite curvature in symmetric spaces? The goal of this talk is to convince the audience that the parameter q, from the point of view of quantization and Kirillov\'s orbit method, is not a quantum parameter but indeed a curvature parameter. The simple relationship is q=e^{-r}, after a modification of the classical definition of quantum groups.
I shall only mention the rank 1 group SL_2, and assume no knowledge of quantum groups. They will have to be (re)defined anyway and discussing the original vision of Drinfeld.
This is joint work with F. Chapon.
15:00 • Université de Genève, Villa Battelle, 7 route de Drize, 1227 Carouge
Prof. Dr. Alessandro Carlotto (ETH Zürich, Switzerland)
Constrained deformations of positive scalar curvature metrics abstract
Abstract:
What manifolds support metrics of positive scalar curvature? If so, what can one say about the associated moduli space? These are two fundamental problems in Riemannian Geometry, for which great progress has been made over the last fifty years, but that are still highly elusive and far from being fully resolved.Partly motivated by the study of initial data sets for the Einstein equations in General Relativity, I will present some results that aim at moving one step further, studying the interplay between two different curvature conditions, given by pointwise inequalities on the scalar curvature of a manifold, and the mean curvature of its boundary. In particular, after a broad contextualization, I will focus on recent joint work with Chao Li (Princeton University), where we give a complete topological characterization of those compact 3-manifolds that support Riemannian metrics of positive scalar curvature and mean-convex boundary and, in any such case, we prove that the associated moduli space of metrics is path-connected. We can also refine our methods so to construct continuous paths of non-negative scalar curvature metrics with minimal boundary, and to obtain analogous conclusions in that context as well. In particular, note that we can derive the path-connectedness of asymptotically flat scalar flat Riemannian 3-manifolds with minimal boundary, which goes in the direction of understanding the space of vacuum black-hole solutions to the Einstein field equations.Our work relies on a combination of earlier fundamental contributions by Gromov-Lawson and Schoen-Yau, on the smoothing procedure designed by Miao to handle singular interfaces, and on the interplay of Perelman\'s Ricci flow with surgery and conformal deformation techniques introduced by Codá Marques in dealing with the closed case.
15:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Chandra Raju (EPF Lausanne)
Title T.B.A.
15:15 • EPF Lausanne, Salle tba
Oliver Gäfvert (KTH, Stockholm)
Parametrised metrics on persistence modules and non-linear hierarchical clustering abstract
Abstract:
In this talk I will show how to construct parametrised classes of metrics on persistence modules facilitated by the introduction of persistence contours. Using persistence contours we can compute the stable rank, an invariant which could previously not be computed, and show that it’s in general NP-hard to compute. Moreover, persistence contours can be used to put persistence modules in a machine learning context by providing a class of metrics that can be optimised over. This could for instance be used to solve classification problems in a metric learning sense, called contour learning.
16:15 • EPF Lausanne, Campus Biotech
Prof. Dr. Peter Bühlmann (ETH Zürich)
Statistical Robustness, Stability and Interpretability of Algorithms
17:15 • UZH Zentrum, Building KO2, Room F 150