DOCSLIB.ORG

Efficient Spectrum-Revealing CUR Matrix Decomposition

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Efficient Spectrum-Revealing CUR Matrix Decomposition Efficient Spectrum-Revealing CUR Matrix Decomposition Cheng Chen Ming Gu Zhihua Zhang Shanghai Jiao Tong University University of California, Berkeley Peking University Weinan Zhang Yong Yu Shanghai Jiao Tong University Shanghai Jiao Tong University Abstract mining, collaborative filtering, bioinformatics, image and video processing (Drineas et al., 2008; Xu et al., The CUR matrix decomposition is an impor- 2015; Wang and Zhang, 2012, 2013; Mahoney et al., tant tool for low-rank matrix approximation. 2008; Mackey et al., 2011). It approximates a data matrix though select- In recent years, many variants of the CUR decom- ing a small number of columns and rows of position have been developed (Drineas et al., 2008; the matrix. Those CUR algorithms with gap- Wang and Zhang, 2012, 2013; Boutsidis and Woodruff, dependent approximation bounds can obtain 2017; Bien et al., 2010; Anderson et al., 2015; Drineas high approximation quality for matrices with et al., 2006; Wang et al., 2016; Song et al., 2019). good singular value spectrum decay, but they Many randomized CUR algorithms adopt random sam- have impractically high time complexities. In pling on columns and rows of data matrices. They this paper, we propose a novel CUR algorithm aim to achieve 1+" relative error bounds on Frobe- based on truncated LU factorization with an nius norm with a failure probability δ (Drineas et al., efficient variant of complete pivoting. Our 2008; Wang and Zhang, 2013; Boutsidis and Woodruff, algorithm has gap-dependent approximation 2 2 2017), i.e., kA − CURkF ≤ (1+")kA − AkkF ; where bounds on both spectral and Frobenius norms k is the target rank of approximation. These bounds while maintaining high efficiency. Numerical are tight theoretically, but are less useful in practi- experiments demonstrate the effectiveness of cal applications because they require O(k=") columns our algorithm and verify our theoretical guar- and rows. For example, Algorithm 3 of (Boutsidis and antees. Woodruff, 2017) requires to select 4k+4820k=" columns and rows, but only gets a success probability of 0:16. If we choose k = 10 and " = 0:5 (very common setting), 1 Introduction the number of selected columns and rows is close to 106. For real-world data sets, it is impractical to select CUR matrix decomposition is a well known method so many columns and rows. Anderson et al. (2015) for low-rank approximation (Mahoney and Drineas, proposed a deterministic column selection method for 2009; Drineas et al., 2008). It approximates the data CUR decomposition with gap-dependent approxima- matrix A as the product of three matrices: A ≈ CUR, tion bounds. These bounds allow their method only to where C and R are composed of sampled columns and choose ` = k+O(1) columns and rows when the given sampled rows of matrix A respectively. Compared matrix has a rapidly decaying spectrum. However, with other low-rank approximation methods such as their algorithm is highly time-consuming and requires truncated Singular Value Decomposition (SVD), CUR O(`mn(m+n)) time, where ` is the number of selected decomposition can better preserve the sparsity of the columns and rows. input matrix and facilitate the interpretation of the computed results. Because of these advantages, CUR A natural question now arises: can we develop an effi- decomposition is more attractive than SVD in text cient CUR algorithm which is as efficient as previous randomized CUR algorithms while maintaining gap- dependent approximation bounds? In this paper, we Proceedings of the 23rdInternational Conference on Artificial address this question via presenting a novel CUR al- Intelligence and Statistics (AISTATS) 2020, Palermo, Italy. gorithm, namely Spectrum-Revealing CUR (SRCUR) PMLR: Volume 108. Copyright 2020 by the author(s). decomposition. Our method borrows the idea from a Efficient Spectrum-Revealing CUR Matrix Decomposition > pP 2 classic numerical linear algebra method, i.e., pivoted For a vector x = (x1; : : : ; xn) , let kxk2 = i xi m×n LU factorization (Trefethen and Bau III, 1997). Unlike be the `2-norm of x. For a matrix A = [aij] 2 R , existing CUR algorithms which sample columns and let kAk2 denote the spectral norm and kAkF denote rows separately (Drineas et al., 2008; Wang and Zhang, the Frobenius norm. We use nnz(A) to denote the 2013; Boutsidis and Woodruff, 2017; Anderson et al., number of nonzero elements of A. Let det(A) be the 2015), the pivoting procedure of LU factorization se- determinant of A and adj(A) be the adjugate matrix lects columns and rows simultaneously and considers of A. We use O~ to hide logarithmic factors. the inter-connectivity of selected columns and rows. Let ρ be the rank of matrix A. The reduced SVD of However, the vanilla pivoting procedure is costly and A is defined as does not bound the relative error. To make use of the ρ advantages and bypass the disadvantages of pivoted > X > LU, we propose Fast Spectrum-Revealing LU (FSRLU) A = UΣV = σiuivi ; factorization, which is efficient and has theoretical guar- i=1 antees. Our SRCUR algorithm first adopts FSRLU to where σi are positive singular values in the descend- select columns and rows and then computes the inter- ing order. That is, σi(A) is the i-th largest singular section matrix. The main contributions of this paper Pk > value of A. Let Ak = i=1 σiuivi denote the best are summarized as follows: rank-k approximation of A and Ay = VΣ−1U> de- note the Moore-Penrose pseudo-inverse of A. κ(A) = • We propose FSRLU algorithm which can efficiently σmax/σmin to denote the condition number of matrix compute truncated LU factorization with complete A. Here σmax is the maximum singular value and σmin pivoting. We also analyze the stability and time is the minimum non-zero singular value of A.of A when complexity of FSRLU. A is square. • We propose SRCUR algorithm based on FSRLU. We use the Matlab colon notation for a block ma- We provide error analysis for SRCUR algorithm trix. Let Ai;: be the i-th row of A and A:;j in both spectral and Frobenius norms. The proofs be the j-th column of A. We use Ai:j;: to de- show that our method has gap dependency error > > > > note Ai;:; Ai+1;:;:::; Aj;: and A:;p:q to denote bounds with regard to the quadratic of the spec- [A ; A ;:::; A ]. Let A denote the block tral gap σ /σ . These error bounds indicate :;p :;p+1 :;q i:j;p:q k+1 p+1 0a : : : a 1 that our method only needs to select ` = k+O(1) ip iq B . .. C columns and rows for data matrices with rapidly matrix @ . A : decaying singular values. ajp : : : ajq • SRCUR is much faster than existing gap- Johnson-Lindenstrauss (JL) transform (Johnson and dependent CUR algorithms. The time complexity Lindenstrauss, 1984) is a powerful tool for dimension- of our method is O(nnz(A) log n)+O~(`2(m+n)), ality reduction. It has been proved to preserve the which is comparable to the state-of-the-art ran- vector norm within an "-error ball. We present the domized CUR algorithms. Johnson-Lindenstrauss Lemma as follows: d • We validate our method with numerical experi- Lemma 1. For any vector x 2 R , 0 < "; δ < 1=2, p×d ments on four real-world datasets, which demon- there exists a JL transform matrix S 2 R , with −2 strate that our algorithms are substantially faster p = Θ(log(1/δ)" ), for which satisfies than existing CUR algorithms while maintaining (1 − ")kxk2 ≤ kSxk2 ≤ (1 + ")kxk2; good performance. 2 2 2 with probability 1 − δ. The remainder of the paper is organized as follows. In Section 2, we describe notation and preliminaries used 3 Related Work in this paper. In Section 3 we introduce related work. Then we present our algorithms in Section 4 and show In Section 3.1, we review some representative methods theoretical analysis in Section 5. Experimental results of CUR matrix decomposition. In Section 3.2, we are given in Section 6. Conclusions and discussions are introduce truncated LU factorization with complete provided in Section 7. pivoting. 2 Notation and Preliminaries 3.1 CUR Matrix Decomposition We use Im to denote m × m identity matrix, and 0 CUR decomposition approximates the data matrix with to denote a zero vector or matrix of appropriate size. actual columns and rows. Given a matrix A 2 Rm×n, Cheng Chen, Ming Gu, Zhihua Zhang, Weinan Zhang, Yong Yu CUR decomposition selects a subset of columns C 2 `1-norm (Song et al., 2017) are less relevant to our Rm×c and a subset of rows R 2 Rr×n and computes an work. intersection matrix U = CyARy. Then, Ab = CUR is a low-rank approximation of matrix A. Some works 3.2 LU factorization with Complete Pivoting try to compute an approximation Ab k with exact rank k (Boutsidis and Woodruff, 2017; Anderson et al., 2015). LU factorization factors a matrix as a product of a The rank-constrained approximation is more compara- unit lower triangular matrix and an upper triangular ble to the truncated SVD decomposition. matrix with row permutation and/or column permu- The essential of CUR decomposition is how to select tation. Partial pivoting is widely used in numerical columns and rows effectively and efficiently. The tra- algebra area because it is more efficient. However, it ditional approaches select columns according to de- may fail on rank deficient matrices. The complete terministic pivoting rules (Stewart, 1999; Chan, 1987; pivoting, even though very expensive, is necessary for Gu and Eisenstat, 1996; Berry et al., 2005), but these those matrices to guarantee that factorized matrices methods do not have good approximation errors.
Load more

Recommended publications
  • Spectral Gap and Asymptotics for a Family of Cocycles of Perron-Frobenius Operators
    Spectral gap and asymptotics for a family of cocycles of Perron-Frobenius operators by Joseph Anthony Horan MSc, University of Victoria, 2015 BMath, University of Waterloo, 2013 A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY in the Department of Mathematics and Statistics c Joseph Anthony Horan, 2020 University of Victoria All rights reserved. This dissertation may not be reproduced in whole or in part, by photocopying or other means, without the permission of the author. We acknowledge with respect the Lekwungen peoples on whose traditional territory the university stands, and the Songhees, Esquimalt, and WSANE´ C´ peoples whose ¯ historical relationships with the land continue to this day. Spectral gap and asymptotics for a family of cocycles of Perron-Frobenius operators by Joseph Anthony Horan MSc, University of Victoria, 2015 BMath, University of Waterloo, 2013 Supervisory Committee Dr. Christopher Bose, Co-Supervisor (Department of Mathematics and Statistics) Dr. Anthony Quas, Co-Supervisor (Department of Mathematics and Statistics) Dr. Sue Whitesides, Outside Member (Department of Computer Science) ii ABSTRACT At its core, a dynamical system is a set of things and rules for how they change. In the study of dynamical systems, we often ask questions about long-term or average phe- nomena: whether or not there is an equilibrium for the system, and if so, how quickly the system approaches that equilibrium. These questions are more challenging in the non-autonomous (or random) setting, where the rules change over time. The main goal of this dissertation is to develop new tools with which to study random dynamical systems, and demonstrate their application in a non-trivial context.
    [Show full text]
  • Nonlinear Spectral Calculus and Super-Expanders
    NONLINEAR SPECTRAL CALCULUS AND SUPER-EXPANDERS MANOR MENDEL AND ASSAF NAOR Abstract. Nonlinear spectral gaps with respect to uniformly convex normed spaces are shown to satisfy a spectral calculus inequality that establishes their decay along Ces`aroaver- ages. Nonlinear spectral gaps of graphs are also shown to behave sub-multiplicatively under zigzag products. These results yield a combinatorial construction of super-expanders, i.e., a sequence of 3-regular graphs that does not admit a coarse embedding into any uniformly convex normed space. Contents 1. Introduction 2 1.1. Coarse non-embeddability 3 1.2. Absolute spectral gaps 5 1.3. A combinatorial approach to the existence of super-expanders 5 1.3.1. The iterative strategy 6 1.3.2. The need for a calculus for nonlinear spectral gaps 7 1.3.3. Metric Markov cotype and spectral calculus 8 1.3.4. The base graph 10 1.3.5. Sub-multiplicativity theorems for graph products 11 2. Preliminary results on nonlinear spectral gaps 12 2.1. The trivial bound for general graphs 13 2.2. γ versus γ+ 14 2.3. Edge completion 18 3. Metric Markov cotype implies nonlinear spectral calculus 19 4. An iterative construction of super-expanders 20 5. The heat semigroup on the tail space 26 5.1. Warmup: the tail space of scalar valued functions 28 5.2. Proof of Theorem 5.1 31 5.3. Proof of Theorem 5.2 34 5.4. Inverting the Laplacian on the vector-valued tail space 37 6. Nonlinear spectral gaps in uniformly convex normed spaces 39 6.1.
    [Show full text]
  • Spectral Gap and Definability
    SPECTRAL GAP AND DEFINABILITY ISAAC GOLDBRING ABSTRACT. We relate the notions of spectral gap for unitary representations and subfactors with definability of certain important sets in the corresponding structures. We give several applications of this relationship. CONTENTS 1. Introduction 2 1.1. A crash course in continuous logic 3 2. Definability in continuous logic 5 2.1. Generalities on formulae 5 2.2. Definability relative to a theory 7 2.3. Definability in a structure 9 3. Spectral gap for unitary group representations 11 3.1. Generalities on unitary group representations 11 3.2. Introducing spectral gap 12 3.3. Spectral gap and definability 13 3.4. Spectral gap and ergodic theory 15 4. Basic von Neumann algebra theory 17 4.1. Preliminaries 17 4.2. Tracial von Neumann algebras as metric structures 19 4.3. Property Gamma and the McDuff property 20 5. Spectral gap subalgebras 21 5.1. Introducing spectral gap for subalgebras 21 arXiv:1805.02752v1 [math.LO] 7 May 2018 5.2. Spectral gap and definability 22 5.3. Relative bicommutants and e.c. II1 factors 23 5.4. Open questions 24 6. Continuum many theories of II1 factors 25 6.1. The history and the main theorem 25 6.2. The base case 26 6.3. A digression on good unitaries and generalized McDuff factors 27 6.4. The inductive step 28 References 29 Goldbring’s work was partially supported by NSF CAREER grant DMS-1349399. 1 2 ISAAC GOLDBRING 1. INTRODUCTION The notion of definable set is one of (if not the) most important concepts in clas- sical model theory.
    [Show full text]
  • Spectral Perturbation Bounds for Selfadjoint Operators I∗
    Spectral perturbation bounds for selfadjoint operators I∗ KreˇsimirVeseli´c† Abstract We give general spectral and eigenvalue perturbation bounds for a selfadjoint op- erator perturbed in the sense of the pseudo-Friedrichs extension. We also give sev- eral generalisations of the aforementioned extension. The spectral bounds for finite eigenvalues are obtained by using analyticity and monotonicity properties (rather than variational principles) and they are general enough to include eigenvalues in gaps of the essential spectrum. 1 Introduction The main purpose of this paper is to derive spectral and eigenvalue bounds for selfadjoint operators. If a selfadjoint operator H in a Hilbert space H is perturbed into T = H + A (1) with, say, a bounded A then the well-known spectral spectral inclusion holds σ(T ) ⊆ {λ : dist(λ, σ(H)) ≤ kAk} . (2) Here σ denotes the spectrum of a linear operator. (Whenever not otherwise stated we shall follow the notation and the terminology of [3].) If H, A, T are finite Hermitian matrices then (1) implies |µk − λk| ≤ kAk, (3) where µk, λk are the non-increasingly ordered eigenvalues of T,H, respectively. (Here and henceforth we count the eigenvalues together with their multiplicities.) Whereas (2) may be called an upper semicontinuity bound the estimate (3) contains an existence statement: each of the intervals [λk − kAk, λk + kAk] contains ’its own’ µk. Colloquially, bounds like (2) may be called ’one-sided’ and those like (3) ’two-sided’. As it is well-known (3) can be refined to another two-sided bound min σ(A) ≤ µk − λk ≤ max σ(A). (4) In [9] the following ’relative’ two-sided bound was derived |µk − λk| ≤ b|λk|, (5) ∗This work was partly done during the author’s stay at the University of Split, Faculty of Electrotechnical Engineering, Mechanical Engineering and Naval Archtecture while supported by the National Foundation for Science, Higher Education and Technological Development of the Republic of Croatia.
    [Show full text]
  • UCLA Electronic Theses and Dissertations
    UCLA UCLA Electronic Theses and Dissertations Title Spectral Gap Rigidity and Unique Prime Decomposition Permalink https://escholarship.org/uc/item/1j63188h Author Winchester, Adam Jeremiah Publication Date 2012 Peer reviewed|Thesis/dissertation eScholarship.org Powered by the California Digital Library University of California University of California Los Angeles Spectral Gap Rigidity and Unique Prime Decomposition A dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy in Mathematics by Adam Jeremiah Winchester 2012 c Copyright by Adam Jeremiah Winchester 2012 Abstract of the Dissertation Spectral Gap Rigidity and Unique Prime Decomposition by Adam Jeremiah Winchester Doctor of Philosophy in Mathematics University of California, Los Angeles, 2012 Professor Sorin Popa, Chair We use malleable deformations combined with spectral gap rigidity theory, in the framework of Popas deformation/rigidity theory to prove unique tensor product decomposition results for II1 factors arising as tensor product of wreath product factors and free group factors. We also obtain a similar result regarding measure equivalence decomposition of direct products of such groups. ii The dissertation of Adam Jeremiah Winchester is approved. Dimitri Shlyakhtenko Yehuda Shalom Jens Palsberg Sorin Popa, Committee Chair University of California, Los Angeles 2012 iii To All My Friends iv Table of Contents 1 Introduction :::::::::::::::::::::::::::::::::::::: 1 2 Preliminaries ::::::::::::::::::::::::::::::::::::: 8 2.1 Intro to von Neumann Algebras . 8 2.2 von Neumann Subalgebras . 13 2.3 GNS and Bimodules . 14 2.4 The Basic Construction . 17 2.5 S-Malleable Deformations . 18 2.6 Spectral Gap Rigidity . 20 2.7 Intertwining By Bimodules . 20 3 Relative Amenability :::::::::::::::::::::::::::::::: 22 3.1 Definition .
    [Show full text]
  • Eigenvalues in Spectral Gaps of Differential Operators
    J. Spectr. Theory 2 (2012), 293–320 Journal of Spectral Theory DOI 10.4171/JST/30 © European Mathematical Society Eigenvalues in spectral gaps of differential operators Marco Marletta1 and Rob Scheichl Abstract. Spectral problems with band-gap spectra arise in numerous applications, including the study of crystalline structures and the determination of transmitted frequencies in pho- tonic waveguides. Numerical discretization of these problems can yield spurious results, a phenomenon known as spectral pollution. We present a method for calculating eigenvalues in the gaps of self-adjoint operators which avoids spectral pollution. The method perturbs the problem into a dissipative problem in which the eigenvalues to be calculated are lifted out of the convex hull of the essential spectrum, away from the spectral pollution. The method is analysed here in the context of one-dimensional Schrödinger equations on the half line, but is applicable in a much wider variety of contexts, including PDEs, block operator matrices, multiplication operators, and others. Mathematics Subject Classification (2010). 34L, 65L15. Keywords. Eigenvalue, spectral pollution, self-adjoint, dissipative, Schrödinger, spectral gap, spectral band, essential spectrum, discretization, variational method. 1. Introduction In the numerical calculation of the spectrum of a self-adjoint operator, one of the most difficult cases to treat arises when the spectrum has band-gap structure and one wishes to calculate eigenvalues in the spectral gaps above the infimum of the essential spectrum. The reason for this difficulty is that variational methods will generally result in spectral pollution (see, e.g., Rappaz, Sanchez Hubert, Sanchez Palencia and Vassiliev [22]): following discretization, the spectral gaps fill up with eigenvalues of the discrete problem which are so closely spaced that it is impossible to distinguish the spectral bands from the spectral gaps.
    [Show full text]
  • Notes by Fred Naud
    HECKE OPERATORS AND SPECTRAL GAPS ON COMPACT LIE GROUPS. FRED´ ERIC´ NAUD, EXTENDED VERSION OF TALK AT PEYRESQ, MAY 22TH 2016 Abstract. In these notes we review some of the recent progress on questions related to spectral gaps of Hecke operators on compact Lie groups. In particular, we discuss the historical Ruziewicz problem that motivated the discovery of the first examples, followed by the work of Gamburd-Jakobson-Sarnak [5], Bourgain-Gamburd [3],[2] and also the recent paper of Benoist-Saxc´e[1]. 1. Introduction Let G be a compact connected Lie Group whose normalized Haar measure is denoted by m. Let µ be a probability measure on G. On the Hilbert space L2(G; dm), we define a convolution operator by Z Tµ(f)(x) := f(xg)dµ(g): G 2 2 We can check easily that Tµ : L ! L is bounded linear with norm 1. The constant function 1 is an obvious eigenfunction related to the eigenvalue 1. If in addition µ is symmetric i.e. −1 µ(A ) = µ(A) for all Borel set A, then Tµ is self-adjoint. The main purpose of these notes is the following question. Set Z 2 2 L0(G) := f 2 L (G): fdm = 0 : G 2 Clearly L0 is a closed Tµ-invariant space. If we have (ρsp is the spectral radius) ρsp(Tµj 2 ) < 1; L0 then we say that Tµ has a spectral gap. There is an easy case: if µ is symmetric and has a continuous density with respect to Haar, then by Ascoli-Arzuela Tµ is a compact operator and one can check that 1 is a simple eigenvalue while −1 is not in the spectrum.
    [Show full text]
  • Spectral Algorithms
    Spectral Algorithms Ravindran Kannan Santosh Vempala August, 2009 ii Summary. Spectral methods refer to the use of eigenvalues, eigenvectors, sin- gular values and singular vectors. They are widely used in Engineering, Ap- plied Mathematics and Statistics. More recently, spectral methods have found numerous applications in Computer Science to \discrete" as well \continuous" problems. This book describes modern applications of spectral methods, and novel algorithms for estimating spectral parameters. In the first part of the book, we present applications of spectral methods to problems from a variety of topics including combinatorial optimization, learning and clustering. The second part of the book is motivated by efficiency considerations. A fea- ture of many modern applications is the massive amount of input data. While sophisticated algorithms for matrix computations have been developed over a century, a more recent development is algorithms based on \sampling on the fly” from massive matrices. Good estimates of singular values and low rank ap- proximations of the whole matrix can be provably derived from a sample. Our main emphasis in the second part of the book is to present these sampling meth- ods with rigorous error bounds. We also present recent extensions of spectral methods from matrices to tensors and their applications to some combinatorial optimization problems. Contents I Applications 1 1 The Best-Fit Subspace 3 1.1 Singular Value Decomposition . .4 1.2 Algorithms for computing the SVD . .8 1.3 The k-variance problem . .8 1.4 Discussion . 11 2 Mixture Models 13 2.1 Probabilistic separation . 14 2.2 Geometric separation . 14 2.3 Spectral Projection .
    [Show full text]
  • On Hyperboundedness and Spectrum of Markov Operators
    On hyperboundedness and spectrum of Markov operators Laurent Miclo Institut de Math´ematiques de Toulouse Plan of the talk 1 Høegh-Krohn and Simon’s conjecture 2 Higher order Cheeger inequalities in the finite setting 3 An approximation procedure 4 General higher order Cheeger’s inequalities 5 Quantitative links between hyperboundedness and spectrum Reversible Markov operators On a probability space S, S,µ , a self-adjoint operator 2 2 p q M : L µ L µ is said to be Markovian if µ-almost surely, p qÑ p q 2 f L µ , f 0 M f 0 @ P p q ě ñ r sě M 1 1 r s “ So M admits a spectral decomposition: there exists a projection-valued measure El 1 1 such that p qlPr´ , s 1 M l dEl “ 1 ż´ and M is said to be ergodic if 2 f L µ , Mf f f Vect 1 @ P p q “ ñ P p q 2 namely if E1 E1 L µ Vect 1 . p ´ ´qr p qs “ p q Hyperboundedness Stronger requirement: M has a spectral gap if there exists λ 0 2 ą such that E1 E1 λ L µ Vect 1 . Spectral gap: the p ´ ´ qr p qs “ p q supremum of such λ. The Markov operator M is hyperbounded if there exists p 2 such ą that M L2 Lp } } pµqÑ pµq ă `8 Theorem A self-adjoint ergodic and hyperbounded Markovian operator admits a spectral gap. This result was conjectured by Høegh-Krohn and Simon [1972], in the context of Markovian semi-groups.
    [Show full text]
  • Spectral Deformations of One-Dimensional Schrö
    SPECTRAL DEFORMATIONS OF ONE-DIMENSIONAL SCHRODINGER OPERATORS E GESZTESY, B. SIMON* AND G. TESCHL Abstract. We provide a complete spectral characterization of a new method of constructing isospectral (in fact, unitary) deformations of general Schrrdinger operators H = -d2/d.x 2 + V in L2(/l~). Our technique is connected to Dirichlet data, that is, the spectrum of the operator H D on L2((-~,x0)) @ L2((x0, oo)) with a Difichlet boundary condition at x 0. The transformation moves a single eigenvalue of H D and perhaps flips which side of x 0 the eigenvalue lives. On the remainder of the spectrum, the transformation is realized by a unitary operator. For cases such as V(x) ---, ~ as Ixl --" ~, where V is uniquely determined by the spectrum of H and the Dirichlet data, our result implies that the specific Dirichlet data allowed are determined only by the asymptotics as E ~ oo. w Introduction Spectral deformations of Schrrdinger operators in L2(/R), isospectral and cer- tain classes of non-isospectral ones, have attracted a lot of interest over the past three decades due to their prominent role in connection with a variety of top- ics, including the Korteweg-de Vries (KdV) hierarchy, inverse spectral problems, supersymmetric quantum mechanical models, level comparison theorems, etc. In fact, the construction of N-soliton solutions of the KdV hierarchy (and more gener- ally, the construction of solitons relative to reflectionless backgrounds) is a typical example of a non-isospectral deformation of H = -d2/dx 2 in L2(R) since the resulting deformation [1 = -d2/dx 2 + re" acquires an additional point spectrum {A1,..., Au} C (-e~, O) such that o'(H) = or(H) tO {A1,..., AN} (cr(.) abbreviating the spectrum).
    [Show full text]
  • QMATH14: Mathematical Results in Quamtum Physics – Abstracts
    QMath14: Mathematical Results in Quantum Physics Aarhus University, 12–16 August 2019 Programme and abstracts – including posters Programme . .next page Abstracts . page 18 Posters . page 63 http://conferences.au.dk/qmath14/ Last update: 2019-08-13 09:53:41 Foto: Anders Trærum, AU Communication Programme Monday, 12 August 09:00–12:00 Mini course A quantum information travel guide for mathematicians Lecturer: David Pérez García Location: Aud E (1533.103) 12:00–14:00 Lunch Location: MATH Canteen (1536, ground floor) 14:00–15:00 Plenary talk Zeno and the bomb Michael Wolf Chair: Pavel Exner Location: Aud E (1533.103) 15:00–15:30 Coffee break 15:30–16:30 Plenary talk An update on Many-Body Localization Wojciech De Roeck Chair: Pavel Exner Location: Aud E (1533.103) 16:30–17:00 Break 17:00– Welcome reception 1 Tuesday, 13 August 09:30–10:30 Plenary talk Quantized quantum transport in interacting systems Sven Bachmann Chair: Simone Warzel Location: Aud E (1533.103) 10:30–11:00 Break 11:00–12:00 Plenary talk Universal Singularities of Random Matrices Torben Krüger Chair: Simone Warzel Location: Aud E (1533.103) 12:00–14:00 Lunch Location: MATH Canteen (1536, ground floor) 14:00–15:30 Parallel sessions – Main speakers Spectral Theory in Aud F (1534.125) see page 8 Quantum Information in Aud G1 (1532.116) see page 9 Many-Body Systems in Aud G2 (1532.122) see page 10 Random Systems in Koll G3 (1532.218) see page 11 Condensed Matter in Koll G4 (1532.222) see page 12 15:30–16:00 Coffee break 2 16:00–17:55 Parallel sessions – Contributed speakers Spectral
    [Show full text]
  • Introduction to the Transfer Operator Method
    Omri Sarig Introduction to the transfer operator method Winter School on Dynamics Hausdorff Research Institute for Mathematics, Bonn, January 2020 Contents 1 Lecture 1: The transfer operator (60 min) ......................... 3 1.1 Motivation . .3 1.2 Definition, basic properties, and examples . .3 1.3 The transfer operator method . .5 2 Lecture 2: Spectral Gap (60 min) ................................ 7 2.1 Quasi–compactness and spectral gap . .7 2.2 Sufficient conditions for quasi-compactness . .9 2.3 Application to continued fractions. .9 3 Lecture 3: Analytic perturbation theory (60 min) . 13 3.1 Calculus in Banach spaces . 13 3.2 Resolvents and eigenprojections . 14 3.3 Analytic perturbations of operators with spectral gap . 15 4 Lecture 4: Application to the Central Limit Theorem (60 min) . 17 4.1 Spectral gap and the central limit theorem . 17 4.2 Background from probability theory . 17 4.3 The proof of the central limit theorem (Nagaev’s method) . 18 5 Lecture 5 (time permitting): Absence of spectral gap (60 min) . 23 5.1 Absence of spectral gap . 23 5.2 Inducing . 23 5.3 Operator renewal theory . 24 A Supplementary material ........................................ 27 A.1 Conditional expectations and Jensen’s inequality . 27 A.2 Mixing and exactness for the Gauss map . 30 A.3 Hennion’s theorem on quasi-compactness . 32 A.4 The analyticity theorem . 40 A.5 Eigenprojections, “separation of spectrum”, and Kato’s Lemma . 41 A.6 The Berry–Esseen “Smoothing Inequality” . 43 1 Lecture 1 The transfer operator 1.1 Motivation A thought experiment Drop a little bit of ink into a glass of water, and then stir it with a tea spoon.
    [Show full text]