Past

Driven by the need for principled extraction of features from time series, we introduce the iterated-sums signature over any commutative semiring. The case of the tropical semiring is a central, and our motivating, example, as it leads to features of (real-valued) time series that are not easily available using existing signature-type objects.

This is joint work with Kurusch Ebrahimi-Fard (NTNU Trondheim) and Nikolas Tapia (WIAS Berlin).

Distinguishing classes of surfaces in $\mathbb{R}^n$ is a task which arises in many situations. There are many characteristics we can use to solve this classification problem. The Persistent Homology Transform allows us to look at shapes in $\mathbb{R}^n$ from $S^{n-1}$ directions simultaneously, and is a useful tool for surface classification. Using the Julia package DiscretePersistentHomologyTransform, we will look at some example curves in $\mathbb{R}^2$ and examine distinguishing features. 

The Gelfand correspondence between compact Hausdorff spaces and unital C*-algebras justifies the slogan that C*-algebras are to be thought of as "non-commutative topological spaces", and Rieffel's theory of compact quantum metric spaces provides, in the same vein, a non-commutative counterpart to the theory of compact metric spaces. The aim of my talk is to introduce the basics of the theory and explain how the classical Gromov-Hausdorff distance between compact metric spaces can be generalized to the quantum setting. If time permits, I will touch upon some recent results obtained in joint work with Jens Kaad and Thomas Gotfredsen.

C*-algebras associated to etale groupoids appear as a versatile construction in many contexts. For instance, groupoid C*-algebras allow for implementation of natural one-parameter groups of automorphisms obtained from continuous cocycles. This provides a path to quantum statistical mechanical systems, where one studies equilibrium states and ground states. The early characterisations of ground states and equilibrium states for groupoid C*-algebras due to Renault have seen remarkable refinements. It is possible to characterise in great generality all ground states of etale groupoid C*-algebras in terms of a boundary groupoid of the cocycle (joint work with Laca and Neshveyev). The steps in the proof employ important constructions for groupoid C*-algebras due to Renault.

We study the minimum Wasserstein distance from the empirical measure to a space of probability measures satisfying linear constraints. This statistic can naturally be used in a wide range of applications, for example, optimally choosing uncertainty sizes in distributionally robust optimization, optimal regularization, testing fairness, martingality, among many other statistical properties. We will discuss duality results which recover the celebrated Kantorovich-Rubinstein duality when the manifold is sufficiently rich and associated test statistics as the sample size increases. We illustrate how this relaxation can beat the statistical curse of dimensionality often associated to empirical Wasserstein distances.

The talk builds on joint work with S. Ghosh, Y. Kang, K. Murthy, M. Squillante, and N. Si.

High dimensionality of biological data is a crucial element that is in need of different methods to unravel their complexity. The current and rich biomedical material that hospitals generate every other day related to cancer detection can benefit from these new techniques. This is the case of diseases such as Acute Lymphoblastic Leukaemia (ALL), one of the most common cancers in childhood. Its diagnosis is based on high-dimensional flow cytometry tumour data that includes immunophenotypic expressions. Not only the intensity of these markers is meaningful for clinicians, but also the shape of the points clouds generated, being then fundamental to find leukaemic clones. Thus, the mathematics of shape recognition in high dimensions can turn itself as a critical tool for this kind of data. This is why we resort to the use of tools from Topological Data Analysis such as Persistence Homology.

Given that ALL relapse incidence is of almost 20% of its patients, we provide a methodology to shed some light on the shape of flow cytometry data, for both relapsed and non-relapsed patients. This is done so by combining the strength of topological data analysis with the versatility of machine learning techniques. The results obtained show us topological differences between both patient sets, such as the amount of connected components and 1-dimensional loops. By means of the so-called persistence images, and for specially selected immunophenotypic markers, a classification of both cohorts is obtained, highlighting the need of new methods to provide better prognosis. 

In the structure theory of operator algebras, it has been a reliable theme that a classification of interesting classes of objects is usually followed by a classification of group actions on said objects. A good example for this is the complete classification of amenable group actions on injective factors, which complemented the famous work of Connes-Haagerup. On the C*-algebra side, progress in the Elliott classification program has likewise given impulse to the classification of C*-dynamics. Although C*-dynamical systems are not yet understood at a comparable level, there are some sophisticated tools in the literature that yield satisfactory partial results. In this introductory talk I will outline the (known) classification of finite group actions with the Rokhlin property, and in the process highlight some themes that are still relevant in today's state-of-the-art.

Quantum groups, which has been a major overarching theme across various branches of mathematics since late 20th century, appear in many ways. Deformation of compact Lie groups is a particularly fruitful paradigm that sits in the intersection between operator algebraic approach to quantized spaces on the one hand, and more algebraic one arising from study of quantum integrable systems on the other.
On the side of operator algebra, Woronowicz defined the C*-bialgebra representing quantized SU(2) based on his theory of pseudospaces. This gives a (noncommutative) C*-algebra of "continuous functions" on the quantized group SUq(2).
Its algebraic counterpart is the quantized universal enveloping algebra Uq(??2), due to Kulish–Reshetikhin and Sklyanin, coming from a search of algebraic structures on solutions of the Yang-Baxter equation. This is (an essentially unique) deformation of the universal enveloping algebra U(??2) as a Hopf algebra.
These structures are in certain duality, and have far-reaching generalization to compact simple Lie groups like SU(n). The interaction of ideas from both fields led to interesting results beyond original settings of these theories.
In this introductory talk, I will explain the basic quantization scheme underlying this "q-deformation", and basic properties of the associated C*-algebras. As part of more recent and advanced topics, I also want to explain an interesting relation to complex simple Lie groups through the idea of quantum double.

The Riemann hypothesis (RH) is one of the great open problems in mathematics. It arose from the study of prime numbers in an analytic context, and—as often occurs in mathematics—developed analogies in an algebraic setting, leading to the influential Weil conjectures. RH for curves over finite fields was proven in the 1940’s by Weil using algebraic-geometric methods. In this talk, we discuss an alternate proof of this result by Stepanov (and Bombieri), using only elementary properties of polynomials. Over the decades, the proof has been whittled down to a 5 page gem! Time permitting, we also indicate connections to exponential sums and the original RH.
 

This talk is motivated by the following question: "how can one reconstruct the geometry of a state space given a collection of observed time series?" A well-studied technique for metric fusion is Similarity Network Fusion (SNF), which works by mixing random walks. However, SNF behaves poorly in the presence of correlated noise, and always reconstructs an intrinsic metric. We propose a new methodology based on delay embeddings, together with a simple orthogonalization scheme that uses the tangency data contained in delay vectors. This method shows promising results for some synthetic and real-world data. The authors suspect that there is a theorem or two hiding in the background -- wild speculation by audience members is encouraged. 

Topological features as computed via persistent homology offer a non-parametric approach to robustly capture multi-scale connectivity information of complex datasets. This has started to gain attention in various machine learning applications. Conventionally, in topological data analysis, this method has been employed as an immutable feature descriptor in order to characterize topological properties of datasets. In this talk, however, I will explore how topological features can be directly integrated into deep learning architectures. This allows us to impose differentiable topological constraints for preserving the global structure of the data space when learning low-dimensional representations.

The TMD algorithm (Kanari et al. 2018) computes the barcode of a neuron (tree) with respect to the radial or path distance from the soma (root). We are interested in the inverse problem: how to understand the space of trees that are represented by the same barcode. Our tool to study this spaces is the stochastic TNS algorithm (Kanari et al. 2020) which generates trees from a given barcode in a biologically meaningful way. 

I will present some theoretical results on the space of trees that have the same barcode, as well as the effect of adding noise to the barcode. In addition, I will provide a more combinatorial perspective on the space of barcodes, expressed in terms of the symmetric group. I will illustrate these results with experiments based on the TNS.

This is joint work with L. Kanari and K. Hess. 

A bundle of C*-algebras is a collection of algebras continuously parametrised by a topological space. There are (at least) two different definitions in operator algebras that make this intuition precise: Continuous C(X)-algebras provide a flexible analytic point of view, while locally trivial C*-algebra bundles allow a classification via homotopy theory. The section algebra of a bundle in the topological sense is a C(X)-algebra, but the converse is not true. In this talk I will compare these two notions using the classical work of Dixmier and Douady on bundles with fibres isomorphic to the compacts  as a guideline. I will then explain joint work with Marius Dadarlat, in which we showed that the theorems of Dixmier and Douady can be generalized to bundles with fibers isomorphic to stabilized strongly self-absorbing C*-algebras. An important feature of the theory is the appearance of higher analogues of the Dixmier-Douady class.

The signature of a path in Euclidean space resides in the tensor algebra of that space; it is obtained by systematic iterated integration of the components of the given path against one another. This straightforward definition conceals a host of deep theoretical properties and impressive practical consequences. In this talk I will describe the homotopical origins of path signatures, their subsequent application to stochastic analysis, and how they facilitate efficient machine learning in topological data analysis. This last bit is joint work with Ilya Chevyrev and Harald Oberhauser.

Quantum limits are objects describing the limit of quadratic quantities (Af_n,f_n) where (f_n) is a sequence of unit vectors in a Hilbert space and A ranges over an algebra of bounded operators. We will discuss the motivation underlying this notion with some important examples from quantum mechanics and from analysis.

The random initialisation of Artificial Neural Networks (ANN) allows one to describe, in the functional space, the limit of the evolution of ANN when their width tends towards infinity. Within this limit, an ANN is initially a Gaussian process and follows, during learning, a gradient descent convoluted by a kernel called the Neural Tangent Kernel.

Connecting neural networks to the well-established theory of kernel methods allows us to understand the dynamics of neural networks, their generalization capability. In practice, it helps to select appropriate architectural features of the network to be trained. In addition, it provides new tools to address the finite size setting.

Since the seminal work of Penrose, it has been understood that conformal compactifications (or "asymptotic simplicity") is the geometrical framework underlying Bondi-Sachs' description of asymptotically flat space-times as an asymptotic expansion. From this point of view the asymptotic boundary, a.k.a "null-infinity", naturally is a conformal null (i.e degenerate) manifold. In particular, "Weyl rescaling" of null-infinity should be understood as gauge transformations. As far as gravitational waves are concerned, it has been well advertised by Ashtekar that if one works with a fixed representative for the conformal metric, gravitational radiations can be neatly parametrized as a choice of "equivalence class of metric-compatible connections". This nice intrinsic description however amounts to working in a fixed gauge and, what is more, the presence of equivalence class tend to make this point of view tedious to work with.

I will review these well-known facts and show how modern methods in conformal geometry (namely tractor calculus) can be adapted to the degenerate conformal geometry of null-infinity to encode the presence of gravitational waves in a completely geometrical (gauge invariant) way: Ashtekar's (equivalence class of) connections are proved to be in 1-1 correspondence with choices of (genuine) tractor connection, gravitational radiation is invariantly described by the tractor curvature and the degeneracy of gravity vacua correspond to the degeneracy of flat tractor connections. The whole construction is fully geometrical and manifestly conformally invariant.

The concept of path signatures has been widely used in several areas of pure mathematics including in applications to data science. However, we remain unable to answer even the most basic questions about it. For instance, how to fully characterise the set of (untruncated) signatures of bounded variation paths? Can certain norms on signatures be related to the length of a path, like in Fourier isometry? In this talk, we will review some known results, explain the open problems and discuss their difficulties.

Abstract: John Roe was a much admired figure in topology and noncommutative geometry, and the creator of the C*-algebraic approach to coarse geometry. John died in 2018 at the age of 58. My aim in the first part of the lecture will be to explain in very general terms the major themes in John’s work, and illustrate them by presenting one of his best-known results, the partitioned manifold index theorem. After the break I shall describe a later result, about relative eta invariants, that has inspired an area of research that is still very active.

Assumed Knowledge: First part: basic familiarity with C*-algebras, plus a little topology. Second part: basic familiarity with K-theory for C*-algebras.

In the first part of the talk, I will review the basic ideas behind Stein’s method for normal approximation and present a general result which we obtained in arXiv:1706.10251 (joint work with Michel Ledoux and Christian Webb). This result states that for a Gibbs measure, an eigenfunction of the corresponding infinitesimal generator is approximately Gaussian in a sense which will be made precise. In the second part, I will report on several applications in random matrix theory. This includes a proof of Johansson’s central limit theorem for linear statistics of beta-ensembles on \R, as well as an application to circular beta-ensembles in the high temperature regime (based on arXiv:1909.01142, joint work with Adrien Hardy).