Unitary solutions of the Yang-Baxter equation ("R-matrices") play a prominent role in several fields, such as quantum field theory and topological quantum computing, but are difficult to find directly and remain somewhat mysterious. In this talk I want to explain how one can use subfactor techniques to learn something about unitary R-matrices, and a research programme aiming at the classification of unitary R-matrices up to a natural equivalence relation. This talk is based on joint work with Roberto Conti, Ulrich Pennig, and Simon Wood.

Identifying discrete patterns in binary data is an important dimensionality reduction tool in machine learning and data mining. In this paper, we consider the problem of low-rank binary matrix factorisation (BMF) under Boolean arithmetic. Due to the NP-hardness of this problem, most previous attempts rely on heuristic techniques. We formulate the problem as a mixed integer linear program and use a large scale optimisation technique of column generation to solve it without the need of heuristic pattern mining. Our approach focuses on accuracy and on the provision of optimality guarantees. Experimental results on real world datasets demonstrate that our proposed method is effective at producing highly accurate factorisations and improves on the previously available best known results for 16 out of 24 problem instances.

--

A link for this talk will be sent to our mailing list a day or two in advance.  If you are not on the list and wish to be sent a link, please contact @email.

In this talk, we will present an approach to constrained optimisation when the set of constraints is a smooth manifold. This setting is of particular interest in data science applications, as many interesting sets of matrices have a manifold structure. We will show how we may couple classic ideas from differential geometry with modern methods such as autodifferentiation to simplify optimisation problems from spaces with a difficult topology (e.g. problems with orthogonal or fixed-rank constraints) to problems on ℝⁿ where we can use any classical optimisation methods to solve them. We will also show how to use these methods to automatically compute quantities such as the Riemannian gradient and Hessian. We will present the library GeoTorch that allows for putting these kind of constraints within models written in PyTorch by adding just one line to the model. We will also comment on some convergence results if time allows.

A link for this talk will be sent to our mailing list a day or two in advance.  If you are not on the list and wish to be sent a link, please contact @email.

We develop a theory to measure the variance and covariance of probability distributions defined on the nodes of a graph, which takes into account the distance between nodes. Our approach generalizes the usual (co)variance to the setting of weighted graphs and retains many of its intuitive and desired properties. As a particular application, we define the maximum-variance problem on graphs with respect to the effective resistance distance, and characterize the solutions to this problem both numerically and theoretically. We show how the maximum-variance distribution can be interpreted as a core-periphery measure, illustrated by the fact that these distributions are supported on the leaf nodes of tree graphs, low-degree nodes in a configuration-like graph and boundary nodes in random geometric graphs. Our theoretical results are supported by a number of experiments on a network of mathematical concepts, where we use the variance and covariance as analytical tools to study the (co-)occurrence of concepts in scientific papers with respect to the (network) relations between these concepts. Finally, I will draw connections to related notion of assortativity on networks, a network analogue of correlation used to describe how the presence and absence of edges covaries with the properties of nodes.

https://arxiv.org/abs/2008.09155

(joint work with T. Kania, Academy of Sciences of the Czech Republic, Prague)
In this talk we discuss our recent result on the inverse eigenvalue problem for symmetric doubly stochastic matrices. 
Namely, we provide a new sufficient condition for a list of real numbers to be the spectrum of a symmetric doubly stochastic matrix. 
In our construction of such matrices, we employ the eigenvectors of the transition probability matrix of a simple symmetric random walk on the circle. 
We also demonstrate a simple algorithm for generating random doubly stochastic matrices based on our construction. Examples will be provided.

In this new session a speaker tells us about how their area of mathematics can be used in different applications.

In this talk, Yuji Nakatsukasa tells us about how random matrix theory can be used in numerical linear algebra. 

Abstract

Randomized SVD is a topic in numerical linear algebra that draws heavily from random matrix theory. It has become an extremely successful approach for efficiently computing a low-rank approximation of matrices. In particular the paper by Halko, Martinsson, and Tropp (SIREV 2011) contains extensive analysis, and has made it a very popular method. The classical Nystrom method is much faster, but only applicable to positive semidefinite matrices. This work studies a generalization of Nystrom's method applicable to general matrices, and shows that (i) it has near-optimal approximation quality comparable to competing methods, (ii) the computational cost is the near-optimal O(mnlog n+r^3) for a rank-r approximation of dense mxn matrices, and (iii) crucially, it can be implemented in a numerically stable fashion despite the presence of an ill-conditioned pseudoinverse. Numerical experiments illustrate that generalized Nystrom can significantly outperform state-of-the-art methods. In this talk I will highlight the crucial role played by a classical result in random matrix theory, namely the Marchenko-Pastur law, and also briefly mention its other applications in least-squares problems and compressed sensing.

After a review of Batalin-Vilkovisky formalism and homotopy algebras, we discuss how these structures emerge in quantum field theory and gravity. We focus then on the application of these sophisticated mathematical tools to scattering amplitudes (both tree- and loop-level) and to the understanding of the dualities between gauge theories and gravity, highlighting generalizations of old results and presenting new ones.

In this talk, I will give a brief introduction to level-set methods for image analysis. I will then describe an application of level-sets to the construction of filtrations for persistent homology computations. I will present several case studies with various spatial data sets using this construction, including applications to voting, analyzing urban street patterns, and spiderwebs. I will conclude by discussing the types of data which I might imagine such methods to be suitable for analyzing and suggesting a few potential future applications of level-set based computations.

In this talk, I will present the theory of combinatorial multivector fields for finite topological spaces, the main subject of my thesis. The idea of combinatorial vector fields came from Forman and emerged naturally from discrete Morse theory. Lately, Mrozek generalized it to the multivector fields theory for Lefschetz complexes. In our work, we simplified and extended it to the finite topological spaces settings. We developed a combinatorial counterpart for dynamical objects, such as isolated invariant sets, isolating neighbourhoods, Conley index, limit sets, and Morse decomposition. We proved the additivity property of the Conley index and the Morse inequalities. Furthermore, we applied persistence homology to study the evolution and the stability of Morse decomposition. In the last part of the talk, I will show numerical results and potential future directions from a data-analysis perspective.