m

F Nous contacter


0

Documents  60B20 | enregistrements trouvés : 24

O

-A +A

P Q

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Post-edited  On determinants of random matrices
Zeitouni, Ofer (Auteur de la Conférence) | CIRM (Editeur )

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

One of the important "products" of wavelet theory consists in the insight that it is often beneficial to consider sparsity in signal processing applications. In fact, wavelet compression relies on the fact that wavelet expansions of real-world signals and images are usually sparse. Compressive sensing builds on sparsity and tells us that sparse signals (expansions) can be recovered from incomplete linear measurements (samples) efficiently. This finding triggered an enormous research activity in recent years both in signal processing applications as well as their mathematical foundations. The present talk discusses connections of compressive sensing and time-frequency analysis (the sister of wavelet theory). In particular, we give on overview on recent results on compressive sensing with time-frequency structured random matrices.

Keywords: compressive sensing - time-frequency analysis - wavelets - sparsity - random matrices - $\ell_1$-minimization - radar - wireless communications
One of the important "products" of wavelet theory consists in the insight that it is often beneficial to consider sparsity in signal processing applications. In fact, wavelet compression relies on the fact that wavelet expansions of real-world signals and images are usually sparse. Compressive sensing builds on sparsity and tells us that sparse signals (expansions) can be recovered from incomplete linear measurements (samples) efficiently. This ...

94A20 ; 94A08 ; 42C40 ; 60B20 ; 90C25

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Multi angle  Free probability and random matrices
Biane, Philippe (Auteur de la Conférence) | CIRM (Editeur )

I will explain how free probability, which is a theory of independence for non-commutative random variables, can be applied to understand the spectra of various models of random matrices.

15B52 ; 60B20 ; 46L53 ; 46L54

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

We will investigate the form of spatio-temporal correlation functions for integrable models of systems of particles on the line. There are few analytical results for nonlinear systems, and so we start developing intuition from harmonic chains, where steepest descent analysis yields detailed asymptotic behaviour of the correlation functions in a variety of scaling limits. We will introduce integrable nonlinear lattices, explain the integrable solution procedure, as well as computational simulations to see dynamics of correlation functions in action.
We will investigate the form of spatio-temporal correlation functions for integrable models of systems of particles on the line. There are few analytical results for nonlinear systems, and so we start developing intuition from harmonic chains, where steepest descent analysis yields detailed asymptotic behaviour of the correlation functions in a variety of scaling limits. We will introduce integrable nonlinear lattices, explain the integrable ...

60B20 ; 60K35 ; 37K10

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

We will investigate the form of spatio-temporal correlation functions for integrable models of systems of particles on the line. There are few analytical results for nonlinear systems, and so we start developing intuition from harmonic chains, where steepest descent analysis yields detailed asymptotic behaviour of the correlation functions in a variety of scaling limits. We will introduce integrable nonlinear lattices, explain the integrable solution procedure, as well as computational simulations to see dynamics of correlation functions in action.
We will investigate the form of spatio-temporal correlation functions for integrable models of systems of particles on the line. There are few analytical results for nonlinear systems, and so we start developing intuition from harmonic chains, where steepest descent analysis yields detailed asymptotic behaviour of the correlation functions in a variety of scaling limits. We will introduce integrable nonlinear lattices, explain the integrable ...

60B20 ; 60K35 ; 37K10

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

We study the expectation of the matrix of overlaps of left and right eigenvectors in the complex Ginibre ensemble, conditioned on a fixed number of k complex eigenvalues.
The diagonal (k=1) and off-diagonal overlap (k=2) were introduced by Chalker and Mehlig. They provided exact expressions for finite matrix size N, in terms of a large determinant of size proportional to N. In the large-N limit these overlaps were determined on the global scale and heuristic arguments for the local scaling at the origin were given. The topic has seen a rapid development in the recent past. Our contribution is to derive exact determinantal expressions of size k x k in terms of a kernel, valid for finite N and arbitrary k.
It can be expressed as an operator acting on the complex eigenvalue correlation functions and allows us to determine all local correlations in the bulk close to the origin, and at the spectral edge. The methods we use are bi-orthogonal polynomials in the complex plane and the analyticity of the diagonal overlap for general k.
This is joint work with Roger Tribe, Athanasios Tsareas, and Oleg Zaboronski as appeared in arXiv:1903.09016 [math-ph]
We study the expectation of the matrix of overlaps of left and right eigenvectors in the complex Ginibre ensemble, conditioned on a fixed number of k complex eigenvalues.
The diagonal (k=1) and off-diagonal overlap (k=2) were introduced by Chalker and Mehlig. They provided exact expressions for finite matrix size N, in terms of a large determinant of size proportional to N. In the large-N limit these overlaps were determined on the global scale ...

60B20 ; 60G55

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

A fundamental question in random matrix theory is to understand how much the eigenvalues of a random matrix fluctuate.
I will address this question in the context of unitary invariant ensembles, by studying the global rigidity of the eigenvalues, or in other words the maximal deviation of an eigenvalue from its classical location.
Our approach to this question combines extreme value theory of log-correlated stochastic processes, and in particular the theory of multiplicative chaos, with asymptotic analysis of large Hankel determinants with Fisher-Hartwig symbols of various types.
In addition to optimal rigidity estimates, our approach sheds light on the extreme values and on the fractal geometry of the eigenvalue counting function.
The talk will be based on joint work in progress with Benjamin Fahs, Gaultier Lambert, and Christian Webb.
A fundamental question in random matrix theory is to understand how much the eigenvalues of a random matrix fluctuate.
I will address this question in the context of unitary invariant ensembles, by studying the global rigidity of the eigenvalues, or in other words the maximal deviation of an eigenvalue from its classical location.
Our approach to this question combines extreme value theory of log-correlated stochastic processes, and in ...

15B52 ; 60B20

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

For the commonly studied Hermitian random matrix models there exist tridiagonal matrix models with the same eigenvalue distribution and the same spectral measure $v_{n}$ at the vector $e_{1}$. These tridiagonal matrices give recurrence coefficients that can be used to build the family of random polynomials that are orthogonal with respect to νn. A similar bijection between spectral data and recurrence coefficients also holds for the Unitary ensembles. This time in stead of obtaining a tridiagonal matrix you obtain a sequence $\left \{ \alpha _{k} \right \}_{k=0}^{n-1}$ Szegö coefficients. The random orthogonal polynomials that are generated by this process may then be used to study properties of the original eigenvalue process.
These techniques may be used not just in the classical cases, but also in the more general case of $\beta $-ensembles. I will discuss various ways that orthogonal polynomials techniques may be applied including to show convergence of the Circular $\beta $-ensemble to $Sine_{\beta }$. I will finish by discussing a result on the maximum deviation of the counting function of Sineβ from it expected value. This is related to studying the phases of associated random orthogonal polynomials.
For the commonly studied Hermitian random matrix models there exist tridiagonal matrix models with the same eigenvalue distribution and the same spectral measure $v_{n}$ at the vector $e_{1}$. These tridiagonal matrices give recurrence coefficients that can be used to build the family of random polynomials that are orthogonal with respect to νn. A similar bijection between spectral data and recurrence coefficients also holds for the Unitary ...

60B20 ; 15B52

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Multi angle  Integrable systems and spectral curves
Eynard, Bertrand (Auteur de la Conférence) | CIRM (Editeur )

Usually one defines a Tau function Tau(t_1,t_2,...) as a function of a family of times having to obey some equations, like Miwa-Jimbo equations, or Hirota equations.
Here we shall view times as local coordinates in the moduli-space of spectral curves, and define the Tau-function of a spectral curve Tau(S), in an intrinsic way, independent of a choice of coordinates. Deformations are tangent vectors, and the tangent space is isomorphic to the space of cycles (cf Goldman bracket), so that Hamiltonians can be represented by cycles.
All the integrable system formalism can then be represented geometrically in the space of cycles: the Poisson bracket is the intersection, the conserved quantities are periods, Miwa-Jimbo equations and Seiberg-Witten equations are a mere consequence of the definition, Hirota equation is a vanishing monodromy condition, and Virasoro-W constraint are automatically satisfied by our definition, showing that our Tau-function is also a conformal block. Our definition contains KdV, KP multicomponent KP, Hitchin systems, and probably all known classical integrable systems.
Usually one defines a Tau function Tau(t_1,t_2,...) as a function of a family of times having to obey some equations, like Miwa-Jimbo equations, or Hirota equations.
Here we shall view times as local coordinates in the moduli-space of spectral curves, and define the Tau-function of a spectral curve Tau(S), in an intrinsic way, independent of a choice of coordinates. Deformations are tangent vectors, and the tangent space is isomorphic to the ...

60B20 ; 37K20

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Multi angle  Universality in tiling models
Van Moerbeke, Pierre (Auteur de la Conférence) | CIRM (Editeur )

We consider the domino tilings of a large class of Aztec rectangles. For an appropriate scaling limit, we show that, the disordered region consists of roughly two arctic circles connected with a finite number of paths. The statistics of these paths is governed by a kernel, also found in other models (universality). The kernel thus obtained is believed to be a master kernel, from which the kernels, associated with critical points, can all be derived.
We consider the domino tilings of a large class of Aztec rectangles. For an appropriate scaling limit, we show that, the disordered region consists of roughly two arctic circles connected with a finite number of paths. The statistics of these paths is governed by a kernel, also found in other models (universality). The kernel thus obtained is believed to be a master kernel, from which the kernels, associated with critical points, can all be ...

60B20 ; 60D05

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Random band matrices (RBM) are natural intermediate models to study eigenvalue statistics and quantum propagation in disordered systems, since they interpolate between mean-field type Wigner matrices and random Schrodinger operators. In particular, RBM can be used to model the Anderson metal-insulator phase transition (crossover) even in 1d. In this talk we will discuss some recent progress in application of the supersymmetric method (SUSY) and transfer matrix approach to the analysis of local spectral characteristics of some specific types of 1d RBM. Joint project with Maria Shcherbina.
Random band matrices (RBM) are natural intermediate models to study eigenvalue statistics and quantum propagation in disordered systems, since they interpolate between mean-field type Wigner matrices and random Schrodinger operators. In particular, RBM can be used to model the Anderson metal-insulator phase transition (crossover) even in 1d. In this talk we will discuss some recent progress in application of the supersymmetric method ...

60B20 ; 15B52

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

The Spherical Sherrington-Kirkpatrick (SSK) model is defined by the Gibbs measure on a highdimensional sphere with a random Hamiltonian given by a symmetric quadratic function. The free energy at the zero temperature is the same as the largest eigenvalue of the random matrix associated with the quadratic function. Even for the finite temperature, there is a simple relationship between the free energy and the eigenvalues. We will discuss how one can study the fluctuations of the free energy using this relationship and results from random matrix theory. We will also discuss the distribution of the spin sampled from the Gibbs measure.
The Spherical Sherrington-Kirkpatrick (SSK) model is defined by the Gibbs measure on a highdimensional sphere with a random Hamiltonian given by a symmetric quadratic function. The free energy at the zero temperature is the same as the largest eigenvalue of the random matrix associated with the quadratic function. Even for the finite temperature, there is a simple relationship between the free energy and the eigenvalues. We will discuss how one ...

60B20 ; 60K35 ; 82D30

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Asymptotic representation theory deals with representations of groups of growing size. For classical Lie groups there are two distinguished regimes of growth. One of them is related to representations of infinite-dimensional groups, and the other appears in combinatorial and probabilistic questions. In the talk I will discuss differences and similarities between these two settings.

22E45 ; 60B20 ; 05E10 ; 60C05

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

I will discuss polynomials $P_{N}$ of degree $N$ that satisfy non-Hermitian orthogonality conditions with respect to the weight $\frac{\left ( z+1 \right )^{N}\left ( z+a \right )^{N}}{z^{2N}}$ on a contour in the complex plane going around 0. These polynomials reduce to Jacobi polynomials in case a = 1 and then their zeros cluster along an open arc on the unit circle as the degree tends to infinity.
For general a, the polynomials are analyzed by a Riemann-Hilbert problem. It follows that the zeros exhibit an interesting transition for the value of a = 1/9, when the open arc closes to form a closed curve with a density that vanishes quadratically. The transition is described by a Painlevé II transcendent.
The polynomials arise in a lozenge tiling problem of a hexagon with a periodic weighting. The transition in the behavior of zeros corresponds to a tacnode in the tiling problem.
This is joint work in progress with Christophe Charlier, Maurice Duits and Jonatan Lenells and we use ideas that were developed in [2] for matrix valued orthogonal polynomials in connection with a domino tiling problem for the Aztec diamond.
I will discuss polynomials $P_{N}$ of degree $N$ that satisfy non-Hermitian orthogonality conditions with respect to the weight $\frac{\left ( z+1 \right )^{N}\left ( z+a \right )^{N}}{z^{2N}}$ on a contour in the complex plane going around 0. These polynomials reduce to Jacobi polynomials in case a = 1 and then their zeros cluster along an open arc on the unit circle as the degree tends to infinity.
For general a, the polynomials are analyzed ...

05B45 ; 52C20 ; 33C45 ; 60B20

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

The universality properties of the Sine process (corresponding to inverse temperature beta equal to 2) are now well known. More generally, a family of point processes have been introduced by Valko and Virag and shown to be the bulk limit of Gaussian beta ensembles, for any positive beta. They are defined through a one-parameter family of SDEs coupled by a two-dimensional Brownian motion (or more recently as the spectrum of a random operator). Through these descriptions, some properties have been derived by Holcomb, Paquette, Valko, Virag and others but there is still much to understand.
In a work with David Dereudre, Adrien Hardy (Université de Lille) and Thomas Leblé (Courant Institute, New York), we use tools from classical statistical mechanics based on DLR equations to give a completely different description of the Sine beta process and derive some properties, such as rigidity and tolerance.
The universality properties of the Sine process (corresponding to inverse temperature beta equal to 2) are now well known. More generally, a family of point processes have been introduced by Valko and Virag and shown to be the bulk limit of Gaussian beta ensembles, for any positive beta. They are defined through a one-parameter family of SDEs coupled by a two-dimensional Brownian motion (or more recently as the spectrum of a random operator). ...

60B20 ; 60G55

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Multi angle  When J. Ginibre met E. Schrödinger
Bothner, Thomas (Auteur de la Conférence) | CIRM (Editeur )

The real Ginibre ensemble consists of square real matrices whose entries are i.i.d. standard normal random variables. In sharp contrast to the complex and quaternion Ginibre ensemble, real eigenvalues in the real Ginibre ensemble attain positive likelihood. In turn, the spectral radius of a real Ginibe matrix follows a different limiting law for purely real eigenvalues than for non-real ones. Building on previous work by Rider, Sinclair and Poplavskyi, Tribe, Zaboronski, we will show that the limiting distribution of the largest real eigenvalue admits a closed form expression in terms of a distinguished solution to an inverse scattering problem for the Zakharov-Shabat system. This system is directly related to several of the most interesting nonlinear evolution equations in 1+1 dimensions which are solvable by the inverse scattering method, for instance the nonlinear Schr¨odinger equation. The results of this talk are based on the recent preprint arXiv:1808.02419, joint with Jinho Baik.
The real Ginibre ensemble consists of square real matrices whose entries are i.i.d. standard normal random variables. In sharp contrast to the complex and quaternion Ginibre ensemble, real eigenvalues in the real Ginibre ensemble attain positive likelihood. In turn, the spectral radius of a real Ginibe matrix follows a different limiting law for purely real eigenvalues than for non-real ones. Building on previous work by Rider, Sinclair and ...

60B20 ; 45M05 ; 60G70

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

In this talk, we discuss the application of the Yang-Baxter equation for the quantum affine lie algebra $U_{q} \left (\widehat{ {\mathfrak{sl}}_{n+1}} \right )$ to interacting particle systems.
The asymmetric simple exclusion process (ASEP) is a continuous-time Markov process of interacting particles on the integer lattice. We distinguish particles to be either a first class or a second class particle. In particular, the second class particles are blocked in their movement by all other particles, while the first class particles are only blocked by other first class particles. We consider the step initial conditions so that all non-negative integer positions are occupied and all other positions are vacant at time zero. Moreover, we take exactly L second class particles to be located at the very front of the configuration at time zero. Then, using recent results of Tracy-Widom (2017) and Borodin-Wheeler (2018), we compute the asymptotic speed of the leftmost second class particle.
This is joint work with Promit Ghosal (Columbia University) and Ethan Zell (University of Virginia) in arXiv:1903.09615.
In this talk, we discuss the application of the Yang-Baxter equation for the quantum affine lie algebra $U_{q} \left (\widehat{ {\mathfrak{sl}}_{n+1}} \right )$ to interacting particle systems.
The asymmetric simple exclusion process (ASEP) is a continuous-time Markov process of interacting particles on the integer lattice. We distinguish particles to be either a first class or a second class particle. In particular, the second class particles ...

34M50 ; 60B20 ; 34E20

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

The talk concerned with the asymptotic empirical eigenvalue distribution of a non linear random matrix ensemble. More precisely we consider $M=
\frac{1}{m} YY^*$ with $Y=f(WX)$ where W and X are random rectangular matrices with i.i.d. centered entries. The function f is applied pointwise and can be seen as an activation function in (random) neural networks. We compute the asymptotic empirical distribution of this ensemble in the case where W and X have subGaussian tails and f is smooth. This extends a result of [PW17] where the case of Gaussian matrices W and X is considered. We also investigate the same questions in the multi-layer case, regarding neural network applications.
The talk concerned with the asymptotic empirical eigenvalue distribution of a non linear random matrix ensemble. More precisely we consider $M=
\frac{1}{m} YY^*$ with $Y=f(WX)$ where W and X are random rectangular matrices with i.i.d. centered entries. The function f is applied pointwise and can be seen as an activation function in (random) neural networks. We compute the asymptotic empirical distribution of this ensemble in the case where W ...

60B20 ; 15B52

Déposez votre fichier ici pour le déplacer vers cet enregistrement.
Déposez votre fichier ici pour le déplacer vers cet enregistrement.

We give a new expression for the law of the eigenvalues of the discrete Anderson model on the finite interval [0, N], in terms of two random processes starting at both ends of the interval. Using this formula, we deduce that the tail of the eigenvectors behaves approximately like exponential of a Brownian motion with a drift. A similar result has recently been shown by B. Rifkind and B. Virag in the critical case, that is, when the random potential is multiplied by a factor 1/ √N.
We give a new expression for the law of the eigenvalues of the discrete Anderson model on the finite interval [0, N], in terms of two random processes starting at both ends of the interval. Using this formula, we deduce that the tail of the eigenvectors behaves approximately like exponential of a Brownian motion with a drift. A similar result has recently been shown by B. Rifkind and B. Virag in the critical case, that is, when the random ...

60B20 ; 65F15

Z