Prof. Dr. Boris Hasselblatt (Tufts University)
Hyperbolic flows abstract
Abstract:
The first nontrivial example of an ergodic mechanical system was themotion of a free particle in a negatively curved space. The underlying mechanism is hyperbolicity. Smooth ergodic theory has since broadenedits scope well beyond uniformly hyperbolic systems, but these provide rich context for investigations connecting dynamical systems to geometry and topology. The lectures present a panorama of such work for uniformly hyperbolic flows. The course begins with a brisk introduction to topological dynamics and ergodic theory of flows and the study of these both for hyperbolic flows. The emphasis will be on continuous-time systems and sometimes their interplay with discrete-time systems.
The first core component of the course then surveys a broad array of mathematics centered on the invariant foliations central to hyperbolic behavior. This centers on their regularity as well as smooth and geometric rigidity. This has a rich interface with geometry.
The second core component studies constructions of Anosov flows, someof them quite recent, which showcases deep interactions of dynamics with low-dimensional topology.
10:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 19.1
Dr. Thibaut Dumont (University of Jyväskylä)
Growth of the volume cocycle in Euclidean buildings abstract
Abstract:
Euclidean buildings are the analogue of symmetric spaces for p-adicLie groups. For example, the (p + 1)-regular tree is the rank one build-ings attached to SL2 over the eld of p-adic number. These buildings areCAT(0) spaces and come with a natural boundary at innity and horo-spherical coordinates. B. Klingler used the latter to introduce a notionof volume. It is an open problem to determine the growth of this volumecocycle related to the cohomology of p-adic Lie groups. The detailedcase of a regular tree has been solved and will be given in parallel.
10:20 • Université de Fribourg, Chemin du Musée 23, Lonza (Math II) seminar room
Dr. Ulrike Riess (ETHZ-ITS)
Log differentials, smoothness, and deformations
13:30 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Enrico Cecini (Università degli Studi di Genova)
Matroid structures in Graph Signal Processing abstract
Abstract:
I will start by reviewing a setting of rising relevance in some applications, namely the so called Graph Signal Processing, that is the study of spaces of functions defined on the vertexes of a graph, whose properties are induced by the eigensystem of a combinatorial laplacian operator. Within this setting I will show how certain natural notions of independent interest, such as admissible sampling subsets, Poisson-like summation formulae and translation operators, can be interpreted in terms of bases and circuits of a certain class of matroids. Next I will derive the defining properties of such matroids, striving to place them within the theory of Coxeter matroids. (This seminar is a report of an ongoing PhD project.)
14:00 • Université de Fribourg, Chemin du Musée 23, Lonza (Math II) seminar room
Carlos Vela Cabello (Autonomous University of Barcelona)
About Z_{2^s}-Linear Hadarmard Codes
15:00 • UZH Irchel, Winterthurerstrasse 190, Zürich, Building Y27, Room H 28
Boris Hasselblatt (Tufts University)
Contact Anosov flows on 3-manifolds abstract
Abstract:
Foulon surgery produces contact Anosov flows of the same topological typeas those of Handel and Thurston and hence different from any algebraicsystem. On one hand, their larger orbit growth implies rapid orbit growthof any Reeb flow for the same contact structure, and on the other handrecent work by Bishop, Hughes, Vinhage and Yang promises a quantificationof the gap between the Liouville and topological entropies.
15:45 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room G 43
Alessandro Verzasconi (University of Zurich)
An investigation on the Roos\'s biases after the RC4 Key Scheduling Algorithm
16:00 • UZH Irchel, Winterthurerstrasse 190, Zürich, Building Y27, Room H 28
Prof. Dr. Olivier Faugeras (INRIA)
Neural networks do not become asynchronous in the large size limit: there is no propagation of chaos abstract
Abstract:
We have developed a new method for establishing the thermodynamic limit of a network of fully connected rate neurons with correlated, Gaussiandistributed, synaptic weights, and random inputs. The method is based on the formulation of a large deviation principle (LDP) for the probabilitydistribution of the neuronal activity of a sequence of networks of increasing sizes. The motivation for using random connections comes fromthe fact that connections in neural networks are complex, poorly known and heterogeneous. The motivation for introducing correlation is theemphasis in computational modelling of neuroscience that neural networks are modular, and the correlations in the connection distributionreproduce this modularity, unlike in previous work. The limiting probability law is Gaussian and its mean and covariance functions are computed using a very quickly converging fixed point algorithm. Our outstanding new result is that in the thermodynamic limit the network does not become asynchronous, there is no propagation of chaos: neurons remain correlated and the amount of correlation can be described precisely from the correlation between the synaptic weights.
16:15 • UZH Irchel, Winterthurerstrasse 190, Zürich, Building Y27, Room H25
Guillaume Rémy (ENS Paris)
The Fyodorov-Bouchaud formula and Liouville conformal field theory abstract
Abstract:
Starting from the restriction of a 2d Gaussian free field (GFF) to theunit circle one can define a Gaussian multiplicative chaos (GMC) measurewhose density is formally given by the exponential of the GFF. In 2008Fyodorov and Bouchaud conjectured an exact formula for the density of thetotal mass of this GMC. In this talk we will give a rigorous proof of thisformula. Our method is inspired by the technology developed by Kupiainen,Rhodes and Vargas to derive the DOZZ formula in the context of Liouvilleconformal field theory on the Riemann sphere. In our case the keyobservation is that the negative moments of the total mass of GMC on thecircle determine its law and are equal to one-point correlation functionsof Liouville theory in the unit disk. Finally we will discuss applicationsin random matrix theory, asymptotics of the maximum of the GFF, and tailexpansions of GMC.
17:15 • UZH Irchel, Winterthurerstrasse 190, Zürich, Building Y27, Room H 12