Abstract:
Functional data such as space curves, images, and surfaces are seldom sensed directly; curves are often noisily and discretely sampled, images are subject to blurring and deformation, and surfaces are often obtained via indirect linear measurements. Understanding the generation of the observations then requires modelling the act of acquisition, and solving an inverse problem in order to arrive at statistical inferences. This leads to a host of new research questions, from identifiability to trade-offs between statistical optimality and computational feasibility, which also make contact with signal processing and numerical analysis.
08:00 • EPF Lausanne, Room GA 3 21
Peter K. Friz (Technische Universität Berlin)
Rough Analysis and Applications (ONLINE) abstract
Abstract:
We plan to discuss: the algebra of iterated integrals, signatures and expected signatures, rough paths and models; Kolmogorov type criteria; Schauder estimates; p-variation, abstract Riemann-Stieltjes integration (Sewing); rough integration and reconstruction; a paracontrolled view; rough differential equations and flows; rough transport equations; second order rough partial differential equations; rough volatility; a regularity structure view; large deviations and precise asymptotics for rough volatility.A good part of the lecture will follow F-Hairer (2014,2020).
10:15 • ETH Zentrum, Rämistrasse 101, Zürich, Building HG, Room
Paul Melotti (Uni Fribourg)
CANCELLED: Cube flips from statistical mechanics to discrete geometry abstract
Abstract:
In various models of statistical mechanics in dimension two, a key property is the existence of a local transformation of the underlying graph and parameters, such that long range properties remain unchanged. This takes the form of a "cube flip" (or star-triangle, or Yang-Baxter equation). On the other hand, there has been recently an intense line of research concerning canonical embeddings of such models (s-embeddings by Chelkak, circle patterns by Kenyon, Lam, Ramassamy and Russkikh, etc). Through this correspondence, cube flips of the model are conjugated to theorems of planar (or projective) geometry on the embedding, and give rise to discrete integrable structures. I will present a few cases of this correspondence. This is based on joint works in progress with Niklas Affolter, Sanjay Ramassamy and Paul Thévenin.
10:20 • Université de Fribourg, Chemin du Musée 23, Lonza (Math II) seminar room
Paul Biran (ETH Zürich)
The Geometry and Algebra of Lagrangian Cobordism abstract
Abstract:
In this talk we will survey the theory of Lagrangian cobordism and discuss the type of structures arising from it: a geometric viewpoint on derived Fukaya categories, new invariants,as well as metric structures on the space of Lagrangians. Based on joint work with O.Cornea.
14:15 • Université de Neuchâtel, Rue Emile Argand 11, B217
Paul Biran (ETH Zürich)
The Geometry and Algebra of Lagrangian Cobordism abstract
Abstract:
In this talk we will survey the theory of Lagrangian cobordism and discuss the type of structures arising from it: a geometric viewpoint on derived Fukaya categories, new invariants,as well as metric structures on the space of Lagrangians. Based on joint work with O.Cornea.
14:15 • Université de Neuchâtel, Rue Emile Argand 11, B217
Simon Riche (Université de Clermont Auvergne)
CANCELLED: Character formulas for reductive algebraic groups abstract
Abstract:
Consider a connected reductive algebraic group G over an algebraically closed field k of positive characteristic. The simple (algebraic) representations of G can easily be classified in terms of their highest weight, but the computation of their characters (in particular, or their dimensions) remains a central problem in this area. In this talk I will present recent progress on this question, obtained in joint works involving Pramod Achar and Geordie Williamson in particular, which involve the study of another class of G-modules (the "tilting" modules) and of a new basis of the associated affine Hecke algebra, called the p-canonical basis.
15:15 • EPF Lausanne, Salle CM1 113
Alex Oancea (Sorbonne Université)
Poincaré duality for loop spaces
15:45 • Université de Neuchâtel, Rue Emile Argand 11, B217
Alex Oancea (Sorbonne Université)
Poincaré duality for loop spaces
15:45 • Université de Neuchâtel, Rue Emile Argand 11, B217
Prof. Dr. Weinan E (Princeton)
A Mathematical Perspective of Machine Learning abstract
Abstract:
The heart of modern machine learning is the approximation of high dimensional functions. Traditional approaches, such as approximation by piecewise polynomials, wavelets, or other linear combinations of fixed basis functions, suffer from the curse of dimensionality. We will discuss representations and approximations that overcome this difficulty, as well as gradient flows that can be used to find the optimal approximation. We will see that at the continuous level, machine learning can be formulated as a series of reasonably nice variational and PDE-like problems. Modern machine learning models/algorithms, such as the random feature and shallow/deep neural network models, can be viewed as special discretizations of such continuous problems. At the theoretical level, we will present a framework that is suited for analyzing machine learning models and algorithms in high dimension, and present results that are free of the curse of dimensionality. Finally, we will discuss the fundamental reasons that are responsible for the success of modern machine learning, as well as the subtleties and mysteries that still remain to be understood.
Prof. Dr. Didier Lucor (LIMSI-CNRS, Paris-Saclay University)
CANCELLED!
16:15 • UZH Irchel, Winterthurerstrasse 190, Zürich, Building Y27, Room H 46