Introduction to Tensor Networks: from Numerical Analysis to Moduli Spaces and Monoidal Categories (Rik Voorhaar, University of Geneva)
04.10.2021 17:00
Tensor networks are a popular tool in mathematics, physics and chemistry to store and manipulate large tensors (multi-dimensional arrays) in an efficient manner. Tensor networks are higher-order generalizations of matrix decompositions. Besides being a useful tool, they are also interesting mathematical objects. Their study is an interesting blend between algebraic/differential geometry and numerical analysis.
In this talk I will introduce the notion of tensor networks, and explain why they're so useful. I will explain the notation of Penrose diagrams (those familiar with monoidal categories will find this familiar). Then I will describe some of the most popular 'formats' of tensor networks. These formats can be described by smooth manifolds or algebraic varieties, and I will explain why this point of view is useful.
Lieu
Bâtiment: Conseil Général 7-9
Room 1-05, Graduate Seminar
Organisé par
Section de mathématiquesIntervenant-e-s
Rik Voorhaar, University of Genevaentrée libre
Plus d'infos
Contact: missing email