Extreme eigenvalue fluctutations for GUE
|
|
- Geraldine Weaver
- 6 years ago
- Views:
Transcription
1 Extreme eigenvalue fluctutations for GUE C. Donati-Martin 204 Program Women and Mathematics, IAS Introduction andom matrices were introduced in multivariate statistics, in the thirties by Wishart [Wis] and in theoretical physics by Wigner [Wig] in the fifties. Since then, the theory developed in a wide range of mathematics fields and physical mathematics. These lectures give a brief introduction to a well studied model : the Gaussian Unitary ensemble (GUE). The GUE is both a Wigner matrix (independent entries) and a model invariant by unitary conjugation. The Gaussian structure enable to compute explicitely some quantities leading to a complete description of the global and local behavior of the spectrum. In particular, in the asymptotics N where N is the size of the matrix, we shall study the fluctuation of the largest eigenvalue around its deterministic limit and prove a central limit theorem towards the so called Tracy-Widom distribution. 2 The Gaussian unitary ensemble - Definition Let H N be the space of Hermitian matrices of size N i.e. matrices M such that M = M. H N is a real vector space of dimension N 2. Definition 2. X N is a gaussian unitary matrix of size N, variance σ 2 (denoted by GUE(N; σ 2 )) if : - X N H N, - The entries of X N satisfy :
2 X N (i, i), i N, 2X N (j, k); j < k, 2IX N (j, k); j < k are independent and distributed as N(0, σ 2 ). One can give an equivalent definition (exercice): Definition 2.2 The distribution P N,σ 2 of GUE(N; σ 2 ) is given by dp N,σ 2(M) = exp( Z N,σ 2 2σ Tr(M 2 ))dm (2.) 2 where dm denotes the Lebesgue measure on H N given by and Z N,σ 2 dm = N i= dm ii is a normalizing constant. i<j N dm ij dim ij We are interesting in the behavior of GUE(N; σ 2 ) in the asymptotics N. The good normalization to see something at the limit is to take σ 2 =. N In the following, we take σ 2 = and denote P N N, by P N. N emark 2. ) GUE(N; ) is a Wigner matrix (independent entries above N the diagonal). Proposition 2. The distribution P N is invariant under unitary conjugation : if X N is P N distributed, then UX N U is also P N distributed, for all unitary matrix U. Proof: We make the change of variable in (2.) : T U : M U MU, for U unitary. T U is an isometry on H N since : T U (M) 2 = Tr((T U (M)) 2 ) = Tr(U MUU MU) = Tr(M 2 ) = M 2. Thus det(t U ) =. On the other hand, exp( 2σ 2 Tr((U MU) 2 ) = exp( 2σ 2 Tr(M 2 )). We conclude, by the change of variable formula, that : E(f(UX N U )) = E(f(X N )). 2
3 2. Distribution of the eigenvalues of GUE(N, N ) Let X N a random matrix distributed as GUE(N, )) and we denote by N λ (X N ) λ N (X N ) the ranked eigenvalues of X N. Proposition 2.2 The joint distribution of the eigenvalues λ (X N ) λ N (X N ) has a density with respect to Lebesgue measure equal to p N (x) = ZN i<j N (x j x i ) 2 exp( N 2 N x 2 i ) x x N (2.2) i= (x) = i<j N (x j x i ) is called the Vandermonde determinant and equals det(x j i ) i,j N. We refer to Mehta [Me, Chap. 3], Deift [D, Chap. 5], Anderson-Guionnet- Zeitouni [AGZ, Chap. 2] for the proof of this proposition. It relies on the expression of the N 2 components of M in (2.) in terms of the N eigenvalues (x i ) and N(N ) independent parameters (p i ) which parametrize the unitary matrix U in the decomposition M = Udiag(x)U. Heuristically, the term exp( N N 2 i= x2 i ) comes from the exp( N Tr(M 2 )) in P 2 N and the square of the Vandermonde determinant comes from the Jacobian of the map M ((x i ), U) after integration on U on the unitary group. Corollary 2. If f is a bounded function of H N, invariant by the unitary transformations, that is f(m) = f(umu ) for all unitary matrix U then f(m) = f(λ (M),, λ N (M)) is a symmetric function of the eigenvalues and E[f(X N )] = Z N H N f(m) exp( 2 N Tr(M 2 ))dm = ZN x x N f(x,, x N ) = N! Z N N f(x,, x N ) i<j N i<j N (x i x j ) 2 exp( N 2 (x i x j ) 2 exp( N k-point correlation functions of the GUE N x 2 i )d N x i= N x 2 i )d N x Let ρ N a symmetric density distribution on N, considered as the distribution of N particles X i. 3 i=
4 Definition 2.3 Let k N. The k- point correlation functions of (X i ) are defined by N! N,k (x, x k ) = ρ N (x, x N )dx k+ dx N. (2.3) (N k)! N k The correlation functions are, up to a constant, the marginal distributions of ρ N. Heuristically, N,k is the probability of finding a particle at x,... N! a particle at x k. The factor comes from the choice of the k particles (N k)! and the symmetry of ρ N (see the computation below). We have, using the symmetry of ρ N, N N E[ ( + f(x i )] = E[ f(x i ) f(x ik )] i= k=0 i < <i k N = E[ f(x i ) f(x ik )] i < <i k and thus, [ N ] E ( + f(x i )) = i= N k=0 = = k=0 N k=0 N k=0 ( N k k! ) E[f(X ) f(x k )] N! (N k)! E[f(X ) f(x k )] f(x ) f(x k ) N,k (x, x N )dx dx k. k! k (2.4) The correlation functions enables to express probabilistic quantities as: ) The hole probability: Take f(x) = \I where I is a Borel set of. Then, the left-hand side of (2.4) is the probability of having no particles in I. Therefore, P( i, X i I) = N ( ) k k=0 k! I k N,k (x, x k )dx dx k. In particular, for I =]a, + [, N ( ) k P(max X i a) = N,k (x, x k )dx dx k. (2.5) k! [a, [ k k=0 4
5 2) the density of state: E[ N f(x i )] = N N i= f(x) N, (x)dx that is N N,(x)dx represents the expectation of the empirical distribution E[ N N i= δ X i ]. We now compute the correlation functions associated to the symmetric density of the (unordered) eigenvalues of the GUE ρ N (x) = N! Z N i<j N (x i x j ) 2 exp( N 2 N x 2 i ). i= Proposition 2.3 The correlation functions of the eigenvalues of GUE(N, N ) are given by N,k(x, xk) = det(kn(xi, xj)) i,j k (2.6) where the kernel K N is given by where K N (x, y) = exp( N N 4 (x2 + y 2 )) q l (x)q l (y) (2.7) q l (x) = where h l are the Hermite polynomials. l=0 ( ) /4 N 2π 2l l! h l( N/2 x) (2.8) The process of the eigenvalues of GUE is said to be a determinantal process. Proof: Since the value of a determinant does not change if we replace a column by the column + a linear combination of the others, we have that the Vandermonde determinant (x) = det(p j (x i )) if P j denotes a polynomial of degree j with higher coefficient equal to. Let w(x) = exp( N 2 x2 ) and define the orthonormal polynomials q l with respect to w such that: - q l is of degree l, q l (x) = a l x l +... with a l > 0. 5
6 - q l(x)q p (x)w(x)dx = δ pl. (q l ) l also depends on N and is up to a scaling factor the family of Hermite polynomials (to be discussed later). Thus, (x) = C N det(q j (x i )) and, using (det(a)) 2 = det(a) det(a T ), we have: ρ N (x) = ZN w(x i ) (det(q j (x i ))) 2 where i N = ZN i N ( N ) w(x i ) det q l (x i )q l (x j ) l= = ZN det (K N (x i, x j )) i,j N K N (x, y) = w(x) w(y) N q l (x)q l (y) = l= N l=0 i,j N φ l (x)φ l (y) where φ l (x) = w(x)q l (x). The sequence (φ l ) l is orthornormal for the Lebesgue measure dx. From the orthonormality of (φ l ), it is easy to show that the kernel K N satisfies the properties: K N (x, x)dx = N K N (x, y)k N (y, z)dy = K N (x, z). This proves (2.6) for k = N (up to a constant). The general case follows from the Lemma: Lemma 2. Let J N = (J ij ) a matrix of size N of the form J ij = f(x i, x j ) with f satisfying:. f(x, x)dx = C 2. f(x, y)f(y, z)dy = f(x, z) Then, det(j N )dx N = (C N + ) det(j N ) where J N is a matrix of size N obtained from J N by removing the last row and column containing x N. 6
7 Proof of Lemma 2.: Exercice Hint: use the formula giving the determinant : det(j N ) = σ Σ N ɛ(σ) N f(x i, x σ(i) ) where Σ N is the set of permutations on {,..., N} and ɛ stands for the signature of a permutation. Next, consider two cases for σ : σ(n) = N and σ(n) N. In the case of GUE, J = (K N (x i, x j )) satisfies the hypothesis of the lemma with C = N. det(k N (x i, x j )) i,j N dx N = (N N ) det(k N (x i, x j )) i,j N det(k N (x i, x j )) i,j N dx N = (N N 2) det(k N (x i, x j )) i,j N 2. Integrating over all the variables gives: det(k N (x i, x j )) i,j N dx... dx N = N! and therefore, Z N = N!. Integrating over the N k variables dx k+,... dx N gives: i= and det(k N (x i, x j )) i,j N dx... dx N = (N k)! det(k N (x i, x j )) i,j k N! k,n (x,... x k ) = ρ N (x,... x N )dx k+... dx N (N k)! = det(k N (x i, x j )) i,j N dx k+... dx N (N k)! = det(k N (x i, x j )) i,j k This proves (2.6) and (2.7). It remains to determine the polynomials q l. Let h l the Hermite polynomial of degree l defined by: ( ) l d h l (x) = ( ) l e x2 (e x2 ). dx 7
8 These polynomials (see [Sz]) are orthogonal with respect to e x2 dx, h2 l (x)e x2 dx = 2 l l! π and the coefficient of x l in h l is 2 l. Then, it is easy to see that q l given by (2.8) are orthonormal with respect to exp( N 2 x2 )dx. Corollary 2.2 Let µ HN (dx) = E[µ HN (dx)] where µ HN is the spectral distribution of GUE(N, ), then µ N H N (dx) is absolutely continuous with respect to Lebesgue measure with density f N given by: f N (x) = N N,(x, x) = N K N(x, x), x. f N is called the density of state. 2.3 The local regime Let us denote, for I a Borel set of, ν N (I) = #{i N; λ i I} = Nµ HN (I) where λ i are the eigenvalues of GUE(N, ). From Wigner s theorem, as N N, ν N (I) N f I sc(x)dx) a.s. where f sc is the density of the semicircular distribution µ sc. The spacing between eigenvalues is of order /N. In the local regime, we consider an interval I N whose size tends to 0 as N. Two cases have to be considered. a) Inside the bulk: Take I N = [u ε N, u + ε N ] with u such that f sc (u) > 0 that is u ] 2, 2[. Then, ν N (I N ) has the order of a constant for ε N. This suggest to introduce new random variables (renormalisation) N l i by l i λ i = u +, i =,... N. Nf sc (u) The mean spacing between the rescaled eigenvalues l i is. Straightforward computations give: Lemma 2.2 The correlation functions bulk of the distribution of (l,..., l N ) are given in terms of the correlation functions of the (λ i ) by bulk N,k (y,..., y k ) = (Nf sc (u)) k N,k(u + y Nf sc (u),..., u + y k ). (2.9) Nf sc (u) We shall see in the next subsection the asymptotic of the correlation functions bulk (or the kernel K N ). 8
9 b) At the edge of the spectrum: u = 2 (or -2). f sc (u) = 0. ν N ([2 ε N, 2]) = N 2π 2 2 ε 4 x2 dx = N 2π So the normalisation at the edge is ε = correlation functions by: edge N,k (y,..., y k ) = N 2/3 ε (N 2/3 ) k N,k(2 + y N 2/3,..., y y2 dy CNε 3/2. and we define the rescaled y k ). (2.0) N 2/3 From (2.5) and (2.0), P[N 2/3 (λ max 2) a] = N ( ) k k=0 k! [a, [ k edge N,k (x, x k )dx dx k. (2.) where λ max is the maximal eigenvalue of the GUE. The asymptotic of edge will be given in the next section. 2.4 Limit kernel The asymptotic of the correlation functions relies on asymptotic formulas for the orthonormal polynomials q l for l N. We have the following: Proposition 2.4 (Plancherel - otach formulas, [Sz]) Let (h n ) n denote the Hermite polynomials. ) If x = 2n + cos(φ) with ε Φ π ε, exp( x 2 /2)h n (x) = b n (sin(φ)) /2 {sin[( n )(sin(2φ) 2Φ)+3π/4]+O( n )} where b n = 2 n/2+/4 (n!) /2 (πn) /4. 2) If x = 2n /2 n /6 t, t bounded in C, exp( x 2 /2)h n (x) = π /4 2 n/2+/4 (n!) /2 (n) /2 {Ai(t) + O( n )} (2.2) where Ai is Airy s function, that is the solution of the differential equation y = xy with y(x) x + 2 π x /4 exp( 2 3 x3/2 ). 9
10 From these formulas, one can show: Theorem 2. lim N bulk n,k (y,..., y k ) = det(k bulk (y i, y j )) i,j k (2.3) where where K bulk (x, y) = sin(π(x y)) π(x y) (2.4) lim N edge n,k (y,..., y k ) = det(k edge (y i, y j )) i,j k (2.5) K edge (x, y) = Ai(x)Ai (y) Ai (x)ai(y) (x y) (2.6) Sketch of Proof of (2.3): From (2.9), (2.6), we may find the limit of Nf sc (u) K N(u + s Nf sc (u), u + t Nf sc (u) ). We express the kernel K N given by (2.7) thanks to Cristoffel-Darboux formula (see Appendix) K N (X, Y ) = k N k N K N (X, Y ) = q N (X)q N (Y ) q N (Y )q N (X) X Y exp( N 4 (X2 + Y 2 )) 2 N (N )! h N ( N/2 X)h N ( N/2 Y ) h N ( N/2 Y )h N ( N/2 X) π X Y exp( N 4 (X2 + Y 2 )) with k N the highest coefficient in q N. Then, set X = u + s, Y = Nf sc(u) u + t, u = 2 cos(φ). Then, f Nf sc(u) sc(u) = sin(φ) and π x = N/2X = 2N(cos(Φ) + πs 2N sin(φ) ). In order to use Plancherel-otach formulas, we express x as x = 2N + cos(φ N ). 0
11 A development gives Φ N = Φ + with a = 2 tan(φ) πs sin 2 (Φ). Then, and a 2N + O( N ) 2 sin(2φ N ) 2Φ N = (sin(2φ) 2Φ) + a (cos(2φ) ) + O( N N ) 2 Formula (2.2) gives: (sin(φ N )) /2 = (sin(φ)) /2 ( + O(/N)). exp( x 2 /2)h N (x) = b N (sin(φ)) /2 {sin[( N )(sin(2φ) 2Φ)+a 2 (cos(2φ) )+3π 4 ]+O(/N)} We make the same transformations for e x2 /2 h N (x), e y2 /2 h N (y), e y2 /2 h N (y) giving φ N, Ψ N and Ψ N associated respectively to: a = 2 tan(φ) πs sin 2 (Φ), b = 2 tan(φ) πt sin 2 (Φ), b = 2 tan(φ) πt sin 2 (Φ). Then, we replace in the product h N (x)h N (y) the product of two sinus by a trigonometric formula and then in the difference, we obtain a linear combination of cosinus, The difference of two of them cancels using that a + b = a + b. Then, we use again a trigonometric formula. After some computations, the kernel K bulk appears. The Airy kernel appears, using the second formula of Plancherel-otach. Corollary 2.3 (Fluctuations of λ max ) The fluctuations of the largest eigenvalue of the GUE around 2 are given by: P(N 2/3 (λ max 2) x) = F 2 (x) where F 2 is called the Tracy-Widom distribution and is given by F 2 (x) = ( ) k k=0 k! ]x, [ k det(k edge (y i, y j )) i,j k d k y. F 2 can be written F 2 (x) = det(i K) L 2 (x, ) where K is the integral operator on L 2 with kernel K edge (x, y) and the det is the Fredholm determinant.
12 2.5 Comments. The computation of the correlation functions which have a determinantal form is specific to the unitary case and do not hold for the GOE case. 2. We refer to [D], [Me] for others computations involving correlation functions such as the spacing distribution. 3. The Tracy-Widom distribution can also be expressed as ( ) F 2 (x) = exp (y x)q 2 (y)dy) x where q (x) = xq(x) + q 3 (x) with q(x) = Ai(x)( + o()) as x. The function q is called the solution of Painlevé II equation (see [TW]). 4. One of the important ideas of the theory is that of universality. This idea is that the asymptotic distribution of some statistic of the eigenvalues in the local scale does not depend very much on the ensemble (like in the TCL), that is the sine kernel (2.4) or the Airy kernel is universal and appears in other models of Hermitian random matrices. This has been shown for - Hermitian Wigner matrices: Soshnikov [So] (for the edge), Johansson [J] for a particular class of matrices. The universality in the bulk was proved in great generality independently by two teams : Erdos, Schlein, Yau [ESY] and their collaborators on one hand and Tao, Vu [TV] on the other hand. - unitary invariant ensemble of the form P N (dm) = C N exp( N Tr(V (M)))dM for a weight V satisfying some assumptions. See [DKMVZ], [PS]. Note that the GUE corresponds to the quadratic weight V (x) = 2 x2. For example, for the Wishart ensemble (associated to the Laguerre polynomials), we have the same asymptotic kernel as in the GUE, while the density of state is not universal (semicircular for GUE and Marchenko- Pastur distribution for Wishart). The main difficulty for general V is to derive the asymptotics of orthogonal polynomials. This can be done using iemann-hilbert techniques (see [D]). 2
13 3 Appendix 3. Change of variable formula Let U and V two open sets in d, g a C diffeomorphism from U to V. If φ is a measurable function on V, positive or Lebesgue integrable, then φ(g(x)) Dg Dx (x) dx = f(y)dy U where Dg (x) is the Jacobian of g, that is the determinant of the Jacobian Dx matrix ( ) gi x j. i,j d 3.2 Van der Monde determinant ecall the van der Monde determinant: det(x j i ) i,j n = i<j(x j x i ). V 3.3 Orthogonal polynomials (see [D], [Sz]) Let w(x) a positive function on such that x m w(x)dx < for all m 0. On the space of real polynomials P [X], we consider the scalar product (P Q) = P (x)q(x)w(x)dx. Then the orthogonalisation procedure of Schmidt enables to construct of sequence of orthogonal polynomials (p l ): p l is of degree l and p m (x)p n (x)w(x)dx = 0 if m n. We denote by a l the coefficient of x l in p l (x) and d l = p l(x) 2 w(x)dx. Example: If w(x) = exp( x 2 ), the Hermite polynomials h l are orthogonal with a l = 2 l and d l = 2 l l! π. 3
14 Christoffel-Darboux formula: We consider a family of orthonormal polynomials (p l ) (d l = ) for the weight w. We denote by K n the kernel defined by: n K n (x, y) = p l (x)p l (y). l=0 K n is the kernel associated to the orthogonal projection in the space of polynomials of degree less than n. This kernel has a simple expression based upon a three terms recurrence relation between the (p l ): xp n (x) = α n p n+ (x) + β n p n (x) + α n p n (x) for some coefficients α n = an a n+ and β n ( depending on a n and the coefficient b n of x n in p n ). From this relation, one obtains: K n (x, y) = α n p n (x)p n (y) p n (y)p n (x) x y (3.) For the orthonormal polynomials q l defined in (2.8), and α N =. a l (= a N,l ) = 3.4 Fredholm determinant ( ) /4 ( ) l N N 2π l! Let K(x, y) a bounded measurable kernel on a space (X, µ) where µ is a finite measure on X. The Fredholm determinant of K is defined by ( λ) k D(λ) = det(i λk) := + det(k(x i, x j )) i,j k µ(dx )... µ(dx k ). k! X k k= The serie converges for all λ. 4
15 eferences [AGZ] Anderson, Greg W.; Guionnet, A.; Zeitouni,. An introduction to random matrices. Cambridge Studies in Advanced Mathematics, 8. Cambridge University Press, Cambridge, 200. [D] Deift, P.A. Orthogonal polynomials and random matrices: a iemann- Hilbert approach. Courant Lecture Notes in Mathematics, 3. New York University, Courant Institute of Mathematical Sciences, New York; American Mathematical Society, Providence, I, 999. [DKMVZ] Deift, P.; Kriecherbauer, T.; McLaughlin, K. T-; Venakides, S.; Zhou, X. Uniform asymptotics for polynomials orthogonal with respect to varying exponential weights and applications to universality questions in random matrix theory. Comm. Pure Appl. Math. 52 (999), no., [ESY] Erdos, L.; Schlein, B.; Yau, H.T. Universality of random matrices and local relaxation flow. Invent. Math. 85 (20), no., 759. [J] Johansson, K. Universality of the local spacing distribution in certain ensembles of Hermitian Wigner matrices. Comm. Math. Phys. 25 (200), no. 3, [Me] Mehta, M.L. andom matrices. Third edition. Pure and Applied Mathematics (Amsterdam), 42. Elsevier/Academic Press, Amsterdam, [PS] Pastur, L. and Shcherbina, M. Universality of the local eigenvalue statistics for a class of unitary invariant random matrix ensembles. J. Statist. Phys. 86 (997), no. -2, [So] Soshnikov, A. Universality at the edge of the spectrum in Wigner random matrices. Comm. Math. Phys. 207 (999), no. 3, [Sz] Szego, G. Orthogonal polynomials. Fourth edition. American Mathematical Society, Colloquium Publications, Vol. XXIII. American Mathematical Society, Providence,.I., 975. [TV] Tao, T.; Vu, V. andom matrices: universality of local eigenvalue statistics. Acta Math. 206 (20), no.,
16 [TW] Tracy, C.A and Widom, H. Level-spacing distributions and the Airy kernel. Comm. Math. Phys. 59 (994), no., [Wig] E. Wigner. On the distribution of the roots of certain symmetric matrices. Ann. of Math. (2) [Wis] J. Wishart The generalized product moment distribution in samples from a normal multivariate population. Biometrika 20 (928),
Determinantal point processes and random matrix theory in a nutshell
Determinantal point processes and random matrix theory in a nutshell part II Manuela Girotti based on M. Girotti s PhD thesis, A. Kuijlaars notes from Les Houches Winter School 202 and B. Eynard s notes
More informationUniversality of distribution functions in random matrix theory Arno Kuijlaars Katholieke Universiteit Leuven, Belgium
Universality of distribution functions in random matrix theory Arno Kuijlaars Katholieke Universiteit Leuven, Belgium SEA 06@MIT, Workshop on Stochastic Eigen-Analysis and its Applications, MIT, Cambridge,
More informationExponential tail inequalities for eigenvalues of random matrices
Exponential tail inequalities for eigenvalues of random matrices M. Ledoux Institut de Mathématiques de Toulouse, France exponential tail inequalities classical theme in probability and statistics quantify
More informationPainlevé Representations for Distribution Functions for Next-Largest, Next-Next-Largest, etc., Eigenvalues of GOE, GUE and GSE
Painlevé Representations for Distribution Functions for Next-Largest, Next-Next-Largest, etc., Eigenvalues of GOE, GUE and GSE Craig A. Tracy UC Davis RHPIA 2005 SISSA, Trieste 1 Figure 1: Paul Painlevé,
More informationMarkov operators, classical orthogonal polynomial ensembles, and random matrices
Markov operators, classical orthogonal polynomial ensembles, and random matrices M. Ledoux, Institut de Mathématiques de Toulouse, France 5ecm Amsterdam, July 2008 recent study of random matrix and random
More informationA Remark on Hypercontractivity and Tail Inequalities for the Largest Eigenvalues of Random Matrices
A Remark on Hypercontractivity and Tail Inequalities for the Largest Eigenvalues of Random Matrices Michel Ledoux Institut de Mathématiques, Université Paul Sabatier, 31062 Toulouse, France E-mail: ledoux@math.ups-tlse.fr
More informationRandom Matrix: From Wigner to Quantum Chaos
Random Matrix: From Wigner to Quantum Chaos Horng-Tzer Yau Harvard University Joint work with P. Bourgade, L. Erdős, B. Schlein and J. Yin 1 Perhaps I am now too courageous when I try to guess the distribution
More informationConcentration Inequalities for Random Matrices
Concentration Inequalities for Random Matrices M. Ledoux Institut de Mathématiques de Toulouse, France exponential tail inequalities classical theme in probability and statistics quantify the asymptotic
More informationDeterminantal point processes and random matrix theory in a nutshell
Determinantal point processes and random matrix theory in a nutshell part I Manuela Girotti based on M. Girotti s PhD thesis and A. Kuijlaars notes from Les Houches Winter School 202 Contents Point Processes
More informationRANDOM MATRIX THEORY AND TOEPLITZ DETERMINANTS
RANDOM MATRIX THEORY AND TOEPLITZ DETERMINANTS David García-García May 13, 2016 Faculdade de Ciências da Universidade de Lisboa OVERVIEW Random Matrix Theory Introduction Matrix ensembles A sample computation:
More informationCentral Limit Theorems for linear statistics for Biorthogonal Ensembles
Central Limit Theorems for linear statistics for Biorthogonal Ensembles Maurice Duits, Stockholm University Based on joint work with Jonathan Breuer (HUJI) Princeton, April 2, 2014 M. Duits (SU) CLT s
More informationUniversality for random matrices and log-gases
Universality for random matrices and log-gases László Erdős IST, Austria Ludwig-Maximilians-Universität, Munich, Germany Encounters Between Discrete and Continuous Mathematics Eötvös Loránd University,
More informationEigenvalue variance bounds for Wigner and covariance random matrices
Eigenvalue variance bounds for Wigner and covariance random matrices S. Dallaporta University of Toulouse, France Abstract. This work is concerned with finite range bounds on the variance of individual
More informationUniversality of local spectral statistics of random matrices
Universality of local spectral statistics of random matrices László Erdős Ludwig-Maximilians-Universität, Munich, Germany CRM, Montreal, Mar 19, 2012 Joint with P. Bourgade, B. Schlein, H.T. Yau, and J.
More informationORTHOGONAL POLYNOMIALS
ORTHOGONAL POLYNOMIALS 1. PRELUDE: THE VAN DER MONDE DETERMINANT The link between random matrix theory and the classical theory of orthogonal polynomials is van der Monde s determinant: 1 1 1 (1) n :=
More informationDEVIATION INEQUALITIES ON LARGEST EIGENVALUES
DEVIATION INEQUALITIES ON LARGEST EIGENVALUES M. Ledoux University of Toulouse, France In these notes, we survey developments on the asymptotic behavior of the largest eigenvalues of random matrix and
More informationarxiv:math/ v2 [math.pr] 10 Aug 2005
arxiv:math/0403090v2 [math.pr] 10 Aug 2005 ORTHOGOAL POLYOMIAL ESEMBLES I PROBABILITY THEORY By Wolfgang König 1 10 August, 2005 Abstract: We survey a number of models from physics, statistical mechanics,
More information1 Intro to RMT (Gene)
M705 Spring 2013 Summary for Week 2 1 Intro to RMT (Gene) (Also see the Anderson - Guionnet - Zeitouni book, pp.6-11(?) ) We start with two independent families of R.V.s, {Z i,j } 1 i
More informationAsymptotics of Hermite polynomials
Asymptotics of Hermite polynomials Michael Lindsey We motivate the study of the asymptotics of Hermite polynomials via their appearance in the analysis of the Gaussian Unitary Ensemble (GUE). Following
More informationThe Density Matrix for the Ground State of 1-d Impenetrable Bosons in a Harmonic Trap
The Density Matrix for the Ground State of 1-d Impenetrable Bosons in a Harmonic Trap Institute of Fundamental Sciences Massey University New Zealand 29 August 2017 A. A. Kapaev Memorial Workshop Michigan
More informationRectangular Young tableaux and the Jacobi ensemble
Rectangular Young tableaux and the Jacobi ensemble Philippe Marchal October 20, 2015 Abstract It has been shown by Pittel and Romik that the random surface associated with a large rectangular Young tableau
More informationLocal semicircle law, Wegner estimate and level repulsion for Wigner random matrices
Local semicircle law, Wegner estimate and level repulsion for Wigner random matrices László Erdős University of Munich Oberwolfach, 2008 Dec Joint work with H.T. Yau (Harvard), B. Schlein (Cambrigde) Goal:
More informationAiry and Pearcey Processes
Airy and Pearcey Processes Craig A. Tracy UC Davis Probability, Geometry and Integrable Systems MSRI December 2005 1 Probability Space: (Ω, Pr, F): Random Matrix Models Gaussian Orthogonal Ensemble (GOE,
More informationOn the concentration of eigenvalues of random symmetric matrices
On the concentration of eigenvalues of random symmetric matrices Noga Alon Michael Krivelevich Van H. Vu April 23, 2012 Abstract It is shown that for every 1 s n, the probability that the s-th largest
More informationFluctuations of random tilings and discrete Beta-ensembles
Fluctuations of random tilings and discrete Beta-ensembles Alice Guionnet CRS (E S Lyon) Advances in Mathematics and Theoretical Physics, Roma, 2017 Joint work with A. Borodin, G. Borot, V. Gorin, J.Huang
More informationDuality, Statistical Mechanics and Random Matrices. Bielefeld Lectures
Duality, Statistical Mechanics and Random Matrices Bielefeld Lectures Tom Spencer Institute for Advanced Study Princeton, NJ August 16, 2016 Overview Statistical mechanics motivated by Random Matrix theory
More informationFluctuations from the Semicircle Law Lecture 4
Fluctuations from the Semicircle Law Lecture 4 Ioana Dumitriu University of Washington Women and Math, IAS 2014 May 23, 2014 Ioana Dumitriu (UW) Fluctuations from the Semicircle Law Lecture 4 May 23, 2014
More informationUpdate on the beta ensembles
Update on the beta ensembles Brian Rider Temple University with M. Krishnapur IISC, J. Ramírez Universidad Costa Rica, B. Virág University of Toronto The Tracy-Widom laws Consider a random Hermitian n
More informationHomogenization of the Dyson Brownian Motion
Homogenization of the Dyson Brownian Motion P. Bourgade, joint work with L. Erdős, J. Yin, H.-T. Yau Cincinnati symposium on probability theory and applications, September 2014 Introduction...........
More informationA determinantal formula for the GOE Tracy-Widom distribution
A determinantal formula for the GOE Tracy-Widom distribution Patrik L. Ferrari and Herbert Spohn Technische Universität München Zentrum Mathematik and Physik Department e-mails: ferrari@ma.tum.de, spohn@ma.tum.de
More informationOPSF, Random Matrices and Riemann-Hilbert problems
OPSF, Random Matrices and Riemann-Hilbert problems School on Orthogonal Polynomials in Approximation Theory and Mathematical Physics, ICMAT 23 27 October, 2017 Plan of the course lecture 1: Orthogonal
More informationStatistical Inference and Random Matrices
Statistical Inference and Random Matrices N.S. Witte Institute of Fundamental Sciences Massey University New Zealand 5-12-2017 Joint work with Peter Forrester 6 th Wellington Workshop in Probability and
More informationBeyond the Gaussian universality class
Beyond the Gaussian universality class MSRI/Evans Talk Ivan Corwin (Courant Institute, NYU) September 13, 2010 Outline Part 1: Random growth models Random deposition, ballistic deposition, corner growth
More informationRandom matrices and determinantal processes
Random matrices and determinantal processes Patrik L. Ferrari Zentrum Mathematik Technische Universität München D-85747 Garching 1 Introduction The aim of this work is to explain some connections between
More informationFrom the mesoscopic to microscopic scale in random matrix theory
From the mesoscopic to microscopic scale in random matrix theory (fixed energy universality for random spectra) With L. Erdős, H.-T. Yau, J. Yin Introduction A spacially confined quantum mechanical system
More informationFluctuations of random tilings and discrete Beta-ensembles
Fluctuations of random tilings and discrete Beta-ensembles Alice Guionnet CRS (E S Lyon) Workshop in geometric functional analysis, MSRI, nov. 13 2017 Joint work with A. Borodin, G. Borot, V. Gorin, J.Huang
More informationOrthogonal Polynomial Ensembles
Chater 11 Orthogonal Polynomial Ensembles 11.1 Orthogonal Polynomials of Scalar rgument Let wx) be a weight function on a real interval, or the unit circle, or generally on some curve in the comlex lane.
More informationComparison Method in Random Matrix Theory
Comparison Method in Random Matrix Theory Jun Yin UW-Madison Valparaíso, Chile, July - 2015 Joint work with A. Knowles. 1 Some random matrices Wigner Matrix: H is N N square matrix, H : H ij = H ji, EH
More informationLectures 6 7 : Marchenko-Pastur Law
Fall 2009 MATH 833 Random Matrices B. Valkó Lectures 6 7 : Marchenko-Pastur Law Notes prepared by: A. Ganguly We will now turn our attention to rectangular matrices. Let X = (X 1, X 2,..., X n ) R p n
More informationFredholm determinant with the confluent hypergeometric kernel
Fredholm determinant with the confluent hypergeometric kernel J. Vasylevska joint work with I. Krasovsky Brunel University Dec 19, 2008 Problem Statement Let K s be the integral operator acting on L 2
More informationLecture I: Asymptotics for large GUE random matrices
Lecture I: Asymptotics for large GUE random matrices Steen Thorbjørnsen, University of Aarhus andom Matrices Definition. Let (Ω, F, P) be a probability space and let n be a positive integer. Then a random
More informationDifferential Equations for Dyson Processes
Differential Equations for Dyson Processes Joint work with Harold Widom I. Overview We call Dyson process any invariant process on ensembles of matrices in which the entries undergo diffusion. Dyson Brownian
More informationThe circular law. Lewis Memorial Lecture / DIMACS minicourse March 19, Terence Tao (UCLA)
The circular law Lewis Memorial Lecture / DIMACS minicourse March 19, 2008 Terence Tao (UCLA) 1 Eigenvalue distributions Let M = (a ij ) 1 i n;1 j n be a square matrix. Then one has n (generalised) eigenvalues
More informationMATH 423 Linear Algebra II Lecture 33: Diagonalization of normal operators.
MATH 423 Linear Algebra II Lecture 33: Diagonalization of normal operators. Adjoint operator and adjoint matrix Given a linear operator L on an inner product space V, the adjoint of L is a transformation
More informationSemicircle law on short scales and delocalization for Wigner random matrices
Semicircle law on short scales and delocalization for Wigner random matrices László Erdős University of Munich Weizmann Institute, December 2007 Joint work with H.T. Yau (Harvard), B. Schlein (Munich)
More informationA Generalization of Wigner s Law
A Generalization of Wigner s Law Inna Zakharevich June 2, 2005 Abstract We present a generalization of Wigner s semicircle law: we consider a sequence of probability distributions (p, p 2,... ), with mean
More informationDynamical approach to random matrix theory
Dynamical approach to random matrix theory László Erdős, Horng-Tzer Yau May 9, 207 Partially supported by ERC Advanced Grant, RAMAT 338804 Partially supported by the SF grant DMS-307444 and a Simons Investigator
More informationCOMPLEX HERMITE POLYNOMIALS: FROM THE SEMI-CIRCULAR LAW TO THE CIRCULAR LAW
Serials Publications www.serialspublications.com OMPLEX HERMITE POLYOMIALS: FROM THE SEMI-IRULAR LAW TO THE IRULAR LAW MIHEL LEDOUX Abstract. We study asymptotics of orthogonal polynomial measures of the
More informationA Note on the Central Limit Theorem for the Eigenvalue Counting Function of Wigner and Covariance Matrices
A Note on the Central Limit Theorem for the Eigenvalue Counting Function of Wigner and Covariance Matrices S. Dallaporta University of Toulouse, France Abstract. This note presents some central limit theorems
More informationRandom Matrices and Multivariate Statistical Analysis
Random Matrices and Multivariate Statistical Analysis Iain Johnstone, Statistics, Stanford imj@stanford.edu SEA 06@MIT p.1 Agenda Classical multivariate techniques Principal Component Analysis Canonical
More informationOPSF, Random Matrices and Riemann-Hilbert problems
OPSF, Random Matrices and Riemann-Hilbert problems School on Orthogonal Polynomials in Approximation Theory and Mathematical Physics, ICMAT 23 27 October, 207 Plan of the course lecture : Orthogonal Polynomials
More informationThe Matrix Dyson Equation in random matrix theory
The Matrix Dyson Equation in random matrix theory László Erdős IST, Austria Mathematical Physics seminar University of Bristol, Feb 3, 207 Joint work with O. Ajanki, T. Krüger Partially supported by ERC
More informationON THE SECOND-ORDER CORRELATION FUNCTION OF THE CHARACTERISTIC POLYNOMIAL OF A HERMITIAN WIGNER MATRIX
ON THE SECOND-ORDER CORRELATION FUNCTION OF THE CHARACTERISTIC POLYNOMIAL OF A HERMITIAN WIGNER MATRIX F. GÖTZE AND H. KÖSTERS Abstract. We consider the asymptotics of the second-order correlation function
More informationNUMERICAL SOLUTION OF DYSON BROWNIAN MOTION AND A SAMPLING SCHEME FOR INVARIANT MATRIX ENSEMBLES
NUMERICAL SOLUTION OF DYSON BROWNIAN MOTION AND A SAMPLING SCHEME FOR INVARIANT MATRIX ENSEMBLES XINGJIE HELEN LI AND GOVIND MENON Abstract. The Dyson Brownian Motion (DBM) describes the stochastic evolution
More informationRandom Toeplitz Matrices
Arnab Sen University of Minnesota Conference on Limits Theorems in Probability, IISc January 11, 2013 Joint work with Bálint Virág What are Toeplitz matrices? a0 a 1 a 2... a1 a0 a 1... a2 a1 a0... a (n
More informationA Note on Universality of the Distribution of the Largest Eigenvalues in Certain Sample Covariance Matrices
Journal of Statistical Physics, Vol. 18, Nos. 5/6, September ( ) A Note on Universality of the Distribution of the Largest Eigenvalues in Certain Sample Covariance Matrices Alexander Soshnikov 1 Received
More informationLecture Notes on the Matrix Dyson Equation and its Applications for Random Matrices
Lecture otes on the Matrix Dyson Equation and its Applications for Random Matrices László Erdős Institute of Science and Technology, Austria June 9, 207 Abstract These lecture notes are a concise introduction
More informationUniformly Random Lozenge Tilings of Polygons on the Triangular Lattice
Interacting Particle Systems, Growth Models and Random Matrices Workshop Uniformly Random Lozenge Tilings of Polygons on the Triangular Lattice Leonid Petrov Department of Mathematics, Northeastern University,
More informationReferences 167. dx n x2 =2
References 1. G. Akemann, J. Baik, P. Di Francesco (editors), The Oxford Handbook of Random Matrix Theory, Oxford University Press, Oxford, 2011. 2. G. Anderson, A. Guionnet, O. Zeitouni, An Introduction
More informationThe norm of polynomials in large random matrices
The norm of polynomials in large random matrices Camille Mâle, École Normale Supérieure de Lyon, Ph.D. Student under the direction of Alice Guionnet. with a significant contribution by Dimitri Shlyakhtenko.
More informationStochastic Differential Equations Related to Soft-Edge Scaling Limit
Stochastic Differential Equations Related to Soft-Edge Scaling Limit Hideki Tanemura Chiba univ. (Japan) joint work with Hirofumi Osada (Kyushu Unv.) 2012 March 29 Hideki Tanemura (Chiba univ.) () SDEs
More informationRandom matrices: Distribution of the least singular value (via Property Testing)
Random matrices: Distribution of the least singular value (via Property Testing) Van H. Vu Department of Mathematics Rutgers vanvu@math.rutgers.edu (joint work with T. Tao, UCLA) 1 Let ξ be a real or complex-valued
More informationarxiv: v1 [math.pr] 9 May 2013
ANDOM MATIX THEOY arxiv:305.253v [math.p] 9 May 203 SLAVA KAGIN STATISTICAL LABOATOY, UNIVESITY OF CAMBIDGE V.KAGIN@STATSLAB.CAM.AC.UK ELENA YUDOVINA STATISTICAL LABOATOY, UNIVESITY OF CAMBIDGE E.YUDOVINA@STATSLAB.CAM.AC.UK
More informationNumerical analysis and random matrix theory. Tom Trogdon UC Irvine
Numerical analysis and random matrix theory Tom Trogdon ttrogdon@math.uci.edu UC Irvine Acknowledgements This is joint work with: Percy Deift Govind Menon Sheehan Olver Raj Rao Numerical analysis and random
More informationDeterminantal Processes And The IID Gaussian Power Series
Determinantal Processes And The IID Gaussian Power Series Yuval Peres U.C. Berkeley Talk based on work joint with: J. Ben Hough Manjunath Krishnapur Bálint Virág 1 Samples of translation invariant point
More informationarxiv: v3 [math-ph] 21 Jun 2012
LOCAL MARCHKO-PASTUR LAW AT TH HARD DG OF SAMPL COVARIAC MATRICS CLAUDIO CACCIAPUOTI, AA MALTSV, AD BJAMI SCHLI arxiv:206.730v3 [math-ph] 2 Jun 202 Abstract. Let X be a matrix whose entries are i.i.d.
More informationFree Probability and Random Matrices: from isomorphisms to universality
Free Probability and Random Matrices: from isomorphisms to universality Alice Guionnet MIT TexAMP, November 21, 2014 Joint works with F. Bekerman, Y. Dabrowski, A.Figalli, E. Maurel-Segala, J. Novak, D.
More informationNear extreme eigenvalues and the first gap of Hermitian random matrices
Near extreme eigenvalues and the first gap of Hermitian random matrices Grégory Schehr LPTMS, CNRS-Université Paris-Sud XI Sydney Random Matrix Theory Workshop 13-16 January 2014 Anthony Perret, G. S.,
More information1 Tridiagonal matrices
Lecture Notes: β-ensembles Bálint Virág Notes with Diane Holcomb 1 Tridiagonal matrices Definition 1. Suppose you have a symmetric matrix A, we can define its spectral measure (at the first coordinate
More informationMaximal height of non-intersecting Brownian motions
Maximal height of non-intersecting Brownian motions G. Schehr Laboratoire de Physique Théorique et Modèles Statistiques CNRS-Université Paris Sud-XI, Orsay Collaborators: A. Comtet (LPTMS, Orsay) P. J.
More informationCorrelation Functions, Cluster Functions, and Spacing Distributions for Random Matrices
Journal of Statistical Physics, Vol. 92, Nos. 5/6, 1998 Correlation Functions, Cluster Functions, and Spacing Distributions for Random Matrices Craig A. Tracy1 and Harold Widom2 Received April 7, 1998
More informationAPPENDIX A. Background Mathematics. A.1 Linear Algebra. Vector algebra. Let x denote the n-dimensional column vector with components x 1 x 2.
APPENDIX A Background Mathematics A. Linear Algebra A.. Vector algebra Let x denote the n-dimensional column vector with components 0 x x 2 B C @. A x n Definition 6 (scalar product). The scalar product
More informationBulk scaling limits, open questions
Bulk scaling limits, open questions Based on: Continuum limits of random matrices and the Brownian carousel B. Valkó, B. Virág. Inventiones (2009). Eigenvalue statistics for CMV matrices: from Poisson
More informationJINHO BAIK Department of Mathematics University of Michigan Ann Arbor, MI USA
LIMITING DISTRIBUTION OF LAST PASSAGE PERCOLATION MODELS JINHO BAIK Department of Mathematics University of Michigan Ann Arbor, MI 48109 USA E-mail: baik@umich.edu We survey some results and applications
More informationTriangular matrices and biorthogonal ensembles
/26 Triangular matrices and biorthogonal ensembles Dimitris Cheliotis Department of Mathematics University of Athens UK Easter Probability Meeting April 8, 206 2/26 Special densities on R n Example. n
More informationRandom regular digraphs: singularity and spectrum
Random regular digraphs: singularity and spectrum Nick Cook, UCLA Probability Seminar, Stanford University November 2, 2015 Universality Circular law Singularity probability Talk outline 1 Universality
More informationLinear Algebra and Dirac Notation, Pt. 2
Linear Algebra and Dirac Notation, Pt. 2 PHYS 500 - Southern Illinois University February 1, 2017 PHYS 500 - Southern Illinois University Linear Algebra and Dirac Notation, Pt. 2 February 1, 2017 1 / 14
More informationEigenvalues and Singular Values of Random Matrices: A Tutorial Introduction
Random Matrix Theory and its applications to Statistics and Wireless Communications Eigenvalues and Singular Values of Random Matrices: A Tutorial Introduction Sergio Verdú Princeton University National
More informationFree Meixner distributions and random matrices
Free Meixner distributions and random matrices Michael Anshelevich July 13, 2006 Some common distributions first... 1 Gaussian Negative binomial Gamma Pascal chi-square geometric exponential 1 2πt e x2
More informationFluctuations of the free energy of spherical spin glass
Fluctuations of the free energy of spherical spin glass Jinho Baik University of Michigan 2015 May, IMS, Singapore Joint work with Ji Oon Lee (KAIST, Korea) Random matrix theory Real N N Wigner matrix
More informationarxiv:hep-th/ v1 14 Oct 1992
ITD 92/93 11 Level-Spacing Distributions and the Airy Kernel Craig A. Tracy Department of Mathematics and Institute of Theoretical Dynamics, University of California, Davis, CA 95616, USA arxiv:hep-th/9210074v1
More informationLarge Deviations, Linear Statistics, and Scaling Limits for Mahler Ensemble of Complex Random Polynomials
Large Deviations, Linear Statistics, and Scaling Limits for Mahler Ensemble of Complex Random Polynomials Maxim L. Yattselev joint work with Christopher D. Sinclair International Conference on Approximation
More informationRigidity of the 3D hierarchical Coulomb gas. Sourav Chatterjee
Rigidity of point processes Let P be a Poisson point process of intensity n in R d, and let A R d be a set of nonzero volume. Let N(A) := A P. Then E(N(A)) = Var(N(A)) = vol(a)n. Thus, N(A) has fluctuations
More informationThe Tracy-Widom distribution is not infinitely divisible.
The Tracy-Widom distribution is not infinitely divisible. arxiv:1601.02898v1 [math.pr] 12 Jan 2016 J. Armando Domínguez-Molina Facultad de Ciencias Físico-Matemáticas Universidad Autónoma de Sinaloa, México
More informationo f P r o b a b i l i t y Vol. 9 (2004), Paper no. 7, pages Journal URL ejpecp/
E l e c t r o n i c J o u r n a l o f P r o b a b i l i t y Vol. 9 (2004), Paper no. 7, pages 177 208. Journal URL http://www.math.washington.edu/ ejpecp/ DIFFERENTIAL OPERATORS AND SPECTRAL DISTRIBUTIONS
More informationQuantum Diffusion and Delocalization for Random Band Matrices
Quantum Diffusion and Delocalization for Random Band Matrices László Erdős Ludwig-Maximilians-Universität, Munich, Germany Montreal, Mar 22, 2012 Joint with Antti Knowles (Harvard University) 1 INTRODUCTION
More informationClass notes: Approximation
Class notes: Approximation Introduction Vector spaces, linear independence, subspace The goal of Numerical Analysis is to compute approximations We want to approximate eg numbers in R or C vectors in R
More informationStrong Markov property of determinantal processes associated with extended kernels
Strong Markov property of determinantal processes associated with extended kernels Hideki Tanemura Chiba university (Chiba, Japan) (November 22, 2013) Hideki Tanemura (Chiba univ.) () Markov process (November
More informationarxiv: v2 [math.pr] 16 Aug 2014
RANDOM WEIGHTED PROJECTIONS, RANDOM QUADRATIC FORMS AND RANDOM EIGENVECTORS VAN VU DEPARTMENT OF MATHEMATICS, YALE UNIVERSITY arxiv:306.3099v2 [math.pr] 6 Aug 204 KE WANG INSTITUTE FOR MATHEMATICS AND
More informationOrthogonal polynomials with respect to generalized Jacobi measures. Tivadar Danka
Orthogonal polynomials with respect to generalized Jacobi measures Tivadar Danka A thesis submitted for the degree of Doctor of Philosophy Supervisor: Vilmos Totik Doctoral School in Mathematics and Computer
More informationRecurrence Relations and Fast Algorithms
Recurrence Relations and Fast Algorithms Mark Tygert Research Report YALEU/DCS/RR-343 December 29, 2005 Abstract We construct fast algorithms for decomposing into and reconstructing from linear combinations
More informationValerio Cappellini. References
CETER FOR THEORETICAL PHYSICS OF THE POLISH ACADEMY OF SCIECES WARSAW, POLAD RADOM DESITY MATRICES AD THEIR DETERMIATS 4 30 SEPTEMBER 5 TH SFB TR 1 MEETIG OF 006 I PRZEGORZAłY KRAKÓW Valerio Cappellini
More informationNumerical Evaluation of Standard Distributions in Random Matrix Theory
Numerical Evaluation of Standard Distributions in Random Matrix Theory A Review of Folkmar Bornemann s MATLAB Package and Paper Matt Redmond Department of Mathematics Massachusetts Institute of Technology
More informationEcon 2148, fall 2017 Gaussian process priors, reproducing kernel Hilbert spaces, and Splines
Econ 2148, fall 2017 Gaussian process priors, reproducing kernel Hilbert spaces, and Splines Maximilian Kasy Department of Economics, Harvard University 1 / 37 Agenda 6 equivalent representations of the
More informationSEA s workshop- MIT - July 10-14
Matrix-valued Stochastic Processes- Eigenvalues Processes and Free Probability SEA s workshop- MIT - July 10-14 July 13, 2006 Outline Matrix-valued stochastic processes. 1- definition and examples. 2-
More informationSTAT C206A / MATH C223A : Stein s method and applications 1. Lecture 31
STAT C26A / MATH C223A : Stein s method and applications Lecture 3 Lecture date: Nov. 7, 27 Scribe: Anand Sarwate Gaussian concentration recap If W, T ) is a pair of random variables such that for all
More informationHOMEWORK PROBLEMS FROM STRANG S LINEAR ALGEBRA AND ITS APPLICATIONS (4TH EDITION)
HOMEWORK PROBLEMS FROM STRANG S LINEAR ALGEBRA AND ITS APPLICATIONS (4TH EDITION) PROFESSOR STEVEN MILLER: BROWN UNIVERSITY: SPRING 2007 1. CHAPTER 1: MATRICES AND GAUSSIAN ELIMINATION Page 9, # 3: Describe
More informationInvertibility of random matrices
University of Michigan February 2011, Princeton University Origins of Random Matrix Theory Statistics (Wishart matrices) PCA of a multivariate Gaussian distribution. [Gaël Varoquaux s blog gael-varoquaux.info]
More informationLinear Algebra: Matrix Eigenvalue Problems
CHAPTER8 Linear Algebra: Matrix Eigenvalue Problems Chapter 8 p1 A matrix eigenvalue problem considers the vector equation (1) Ax = λx. 8.0 Linear Algebra: Matrix Eigenvalue Problems Here A is a given
More informationLecture 23: Trace and determinants! (1) (Final lecture)
Lecture 23: Trace and determinants! (1) (Final lecture) Travis Schedler Thurs, Dec 9, 2010 (version: Monday, Dec 13, 3:52 PM) Goals (2) Recall χ T (x) = (x λ 1 ) (x λ n ) = x n tr(t )x n 1 + +( 1) n det(t
More information