DOCSLIB.ORG

Arxiv:Quant-Ph/0408115V2 16 Jun 2005 Admresult Random Operational the of Issues Theory[4]

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Arxiv:Quant-Ph/0408115V2 16 Jun 2005 Admresult Random Operational the of Issues Theory[4] Classical randomness in quantum measurements Giacomo Mauro D’Ariano,∗ Paoloplacido Lo Presti,† and Paolo Perinotti‡ QUIT – Quantum Information Theory Group, Dipartimento di Fisica “A. Volta”, via Bassi 6, I-27100 Pavia, Italy (Dated: October 23, 2018) Similarly to quantum states, also quantum measurements can be ”mixed”, corresponding to a random choice within an ensemble of measuring apparatuses. Such mixing is equivalent to a sort of hidden variable, which produces a noise of purely classical nature. It is then natural to ask which apparatuses are indecomposable, i. e. do not correspond to any random choice of apparatuses. This problem is interesting not only for foundations, but also for applications, since most optimization strategies give optimal apparatuses that are indecomposable. Mathematically the problem is posed describing each measuring apparatus by a positive operator- valued measure (POVM), which gives the statistics of the outcomes for any input state. The POVM’s form a convex set, and in this language the indecomposable apparatuses are represented by extremal points—the analogous of ”pure states” in the convex set of states. Differently from the case of states, however, indecomposable POVM’s are not necessarily rank-one, e. g. von Neumann measurements. In this paper we give a complete classification of indecomposable apparatuses (for discrete spec- trum), by providing different necessary and sufficient conditions for extremality of POVM’s, along with a simple general algorithm for the decomposition of a POVM into extremals. As an interesting application, ”informationally complete” measurements are analyzed in this respect. The convex set of POVM’s is fully characterized by determining its border in terms of simple algebraic properties of the corresponding POVM’s. PACS numbers: 03.65.Ta, 03.67.-a I. INTRODUCTION of the measuring apparatus according to the Born rule p(e|ρ) = Tr[ρPe]. (1) Measurements are the essence of any experimental sci- In Eq. (1) Pe denote positive operators on the Hilbert ence. At extreme sensitivities and precisions they become space H of the system, representing our knowledge of the true core of quantum mechanics. For any practical the measuring apparatus from which we infer informa- need, a measurement can be always regarded as the re- tion on the state ρ from the probability distribution trieval of information about the ”state” of the measured p(e|ρ). Positivity of Pe is needed for positivity of p(e|ρ), system. However, due to the no-cloning theorem[1, 2], whereas normalization is guaranteed by the completeness even for an elementary system—e.g. a single harmonic Pe∈E Pe = I. In the present paper we will only consider oscillator or a spin—it is impossible to recover a com- the simple case of finite discrete set E. More generally, plete knowledge of the state of the system from a single on has an infinite probability space E (generally contin- measurement[3] without prior knowledge. Then, since uous), and in this context the set of positive operators in quantum mechanics different incompatible measure- {Pe} becomes actually a positive operator valued mea- ments can be performed in principle, one is faced with sure (POVM). Every apparatus is described by a POVM, the problem of which measurement should be adopted and, reversely, every POVM can be realized in principle arXiv:quant-ph/0408115v2 16 Jun 2005 for accomplishing a specific task, and which strategy of by an apparatus [5, 6, 7, 8, 9, 10, 11]. repeated measurements would be the most statistically The linearity of the Born rule (1) in both arguments ρ efficient. These are the basic issues of the operational and Pe is consistent with the intrinsically statistical na- viewpoint of Quantum Estimation Theory[4]. ture of the measurement, in which our partial knowledge of both the system and the apparatus reflects in ”con- A measurement on a quantum system [5] returns a vex” structures for both states and POVM’s. This means random result e from a set of possible outcomes E = {e = that not only states, but also POVM’s can be ”mixed”, 1,...N}, with probability distribution p(e|ρ) depending namely there are POVM’s that give probability distri- on the state ρ of the system in a way which is distinctive butions p(e|ρ) that are equivalent to randomly choosing among different apparatuses. Notice that mixed POVM’s can also correspond to a single measuring apparatus, not only when the apparatus itself is prepared in a mixed ∗Electronic address: [email protected]; Also at Center for Pho- state, but also for pure preparation, as a result of discard- tonic Communication and Computing, Department of Electrical ing (tracing out) the apparatus after a unitary interaction and Computer Engineering, Northwestern University, Evanston, IL with the system. Clearly, such mixing is itself a source of 60208 †Electronic address: [email protected] ”classical” noise, which can be in principle removed by ‡Electronic address: perinotti@fisicavolta.unipv.it; Also at Istituto adopting an indecomposable apparatus in the ensemble Nazionale di Fisica della Materia Unit`adi Pavia corresponding to the mixed POVM. It is then natural to 2 ask which apparatuses are indecomposable, i. e. ”pure” After summarizing the results in the concluding Section in the above sense, or, mathematically, which POVM’s VI, Appendix A reports the algorithm for decomposing correspond to extremal points of the convex set. The a POVM into extremals. classification of such apparatuses is certainly very use- ful in applications, since most optimization strategies in quantum estimation theory[4, 5] correspond to minimize II. CONVEXITY AND EXTREMALITY OF a concave (actually linear) function on the POVM’s con- POVM’S vex set—so-called ”cost-function”—whence leading to an optimal POVM which is extremal. Let’s denote by PN the convex set of POVM’s on a Surprisingly, extremality for POVM’s generally doesn’t finite dimensional Hilbert space H, with a number N of mean to be rank-one, as for ”pure” states. In other outcomes E = {1,...,N}. We will represent a POVM words, indecomposable POVM’s are not necessarily re- in the set as the vector P ∈ PN P = {P1,...,PN } alized by von Neumann measurements. Indeed, as we of the N positive operators Pe. The fact that the set will see in this paper, there are rank-one POVM’s that PN is convex means that it is closed under convex lin- ′ ′′ are not extremal, whereas, on the opposite side, there ear combinations, namely for any P , P ∈ PN also ′ ′′ are higher-rank POVM’s which are extremal. Moreover, P = pP + (1 − p)P ∈ PN with 0 ≤ p ≤ 1 — i.e. p is whenever the optimization problems have additional lin- a probability. Then, P can be also equivalently achieved ear constraints—e. g. for covariant POVM’s, or for by randomly choosing between two different apparatuses fixed probability distribution on a given state—the cor- corresponding to P′ and P′′, respectively, with probabil- responding subset of POVM’s is a lower dimensional con- ity p and 1 − p, since, the overall statistical distribution vex set corresponding to a section by a hyperplane of the p(e|ρ) will be the convex combination of the statistics ′ ′′ complete POVM’s convex set, with boundary equal to coming from P and P . Notice that PN contains also the section of the original boundary, and whence with the set of POVM’s with a strictly smaller number of out- extremal points that belong to the boundary of the con- comes, i. e. with E′ ⊂ E. For such POVM’s the elements vex set of all POVM’s. For this reason, also the boundary corresponding to outcomes in E \ E′ will be zero, corre- of the POVM’s convex set is interesting in practice, since sponding to zero probability of occurrence for all states. POVM’s that are optimal (for a concave cost function) Clearly, for N ≤ M one has PN ⊆ PM ⊆ P, where by with an additional linear constraint generally are non ex- P we denote the convex set of all POVM’s with any (gen- tremal, but still belong to the boundary. One can also erally infinite) discrete spectrum. The extremal points argue that POVM’s which lie inside faces of the convex, of PN represent the ”indecomposable” measurements, physically exhibit a different degree of ”classical” noise which cannot be achieved by mixing different measure- in relation with the dimensionality of the face. Notice ments. Obviously, a POVM which is extremal for PN is that one should not imagine the POVM’s set as a poly- also extremal for PM with M ≥ N, whence it is actu- tope, since, on the contrary the set is “strongly convex”, ally extremal for P, and we will simply name it extremal namely the extremal points are not isolated, but lie on without further specifications. continuous manifolds. Let’s start with a simple example. Consider the fol- In the present paper we address the problem of appa- lowing two-outcome POVM for a qubit ratus decomposability along three lines of attack: a) by 1 1 providing simple necessary and sufficient conditions for P = 2 |0ih0|, 2 |0ih0| + |1ih1| . (2) extremality of POVM’s; b) by establishing the complete 1 structure of the POVM’s convex set via the character- By defining D = 2 (|0ih0|, −|0ih0|), the two vectors P± = ization of its border in terms of algebraic properties of P ± D correspond to the following different POVM’s the POVM’s; c) by providing a simple general algorithm for the decomposition of POVM’s into extremals. For P+ = (|0ih0|, |1ih1|) , P− = (0, I) , (3) simplicity, the whole paper is restricted to the case of discrete spectrum.
Load more

Recommended publications
  • Ekeland Variational Principles with Set-Valued Objective Functions And
    Ekeland variational principles with set-valued objective functions and set-valued perturbations1 QIU JingHui School of Mathematical Sciences, Soochow University, Suzhou 215006, China Email: [email protected] Abstract In the setting of real vector spaces, we establish a general set-valued Ekeland variational principle (briefly, denoted by EVP), where the objective func- tion is a set-valued map taking values in a real vector space quasi-ordered by a convex cone K and the perturbation consists of a K-convex subset H of the or- dering cone K multiplied by the distance function. Here, the assumption on lower boundedness of the objective function is taken to be the weakest kind. From the general set-valued EVP, we deduce a number of particular versions of set-valued EVP, which extend and improve the related results in the literature. In particu- lar, we give several EVPs for approximately efficient solutions in set-valued opti- mization, where a usual assumption for K-boundedness (by scalarization) of the objective function’s range is removed. Moreover, still under the weakest lower boundedness condition, we present a set-valued EVP, where the objective function is a set-valued map taking values in a quasi-ordered topological vector space and the perturbation consists of a σ-convex subset of the ordering cone multiplied by the distance function. arXiv:1708.05126v1 [math.FA] 17 Aug 2017 Keywords nonconvex separation functional, vector closure, Ekeland variational principle, coradiant set, (C, ǫ)-efficient solution, σ-convex set MSC(2010) 46A03, 49J53, 58E30, 65K10, 90C48 1 Introduction Since the variational principle of Ekeland [12, 13] for approximate solutions of nonconvex minimization problems appeared in 1972, there have been various 1This work was supported by the National Natural Science Foundation of China (Grant Nos.
    [Show full text]
  • Techniques of Variational Analysis
    J. M. Borwein and Q. J. Zhu Techniques of Variational Analysis An Introduction October 8, 2004 Springer Berlin Heidelberg NewYork Hong Kong London Milan Paris Tokyo To Tova, Naomi, Rachel and Judith. To Charles and Lilly. And in fond and respectful memory of Simon Fitzpatrick (1953-2004). Preface Variational arguments are classical techniques whose use can be traced back to the early development of the calculus of variations and further. Rooted in the physical principle of least action they have wide applications in diverse ¯elds. The discovery of modern variational principles and nonsmooth analysis further expand the range of applications of these techniques. The motivation to write this book came from a desire to share our pleasure in applying such variational techniques and promoting these powerful tools. Potential readers of this book will be researchers and graduate students who might bene¯t from using variational methods. The only broad prerequisite we anticipate is a working knowledge of un- dergraduate analysis and of the basic principles of functional analysis (e.g., those encountered in a typical introductory functional analysis course). We hope to attract researchers from diverse areas { who may fruitfully use varia- tional techniques { by providing them with a relatively systematical account of the principles of variational analysis. We also hope to give further insight to graduate students whose research already concentrates on variational analysis. Keeping these two di®erent reader groups in mind we arrange the material into relatively independent blocks. We discuss various forms of variational princi- ples early in Chapter 2. We then discuss applications of variational techniques in di®erent areas in Chapters 3{7.
    [Show full text]
  • On the Convexity of Optimal Control Problems Involving Non-Linear Pdes Or Vis and Applications to Nash Games
    Weierstraß-Institut für Angewandte Analysis und Stochastik Leibniz-Institut im Forschungsverbund Berlin e. V. Preprint ISSN 2198-5855 On the convexity of optimal control problems involving non-linear PDEs or VIs and applications to Nash games Michael Hintermüller1,2, Steven-Marian Stengl1,2 submitted: September 11, 2020 1 Weierstrass Institute 2 Humboldt-Universität zu Berlin Mohrenstr. 39 Unter den Linden 6 10117 Berlin 10099 Berlin Germany Germany E-Mail: [email protected] E-Mail: [email protected] [email protected] [email protected] No. 2759 Berlin 2020 2010 Mathematics Subject Classification. 06B99, 49K21, 47H04, 49J40. Key words and phrases. PDE-constrained optimization, K-Convexity, set-valued analysis, subdifferential, semilinear elliptic PDEs, variational inequality. This research was supported by the DFG under the grant HI 1466/10-1 associated to the project “Generalized Nash Equilibrium Problems with Partial Differential Operators: Theory, Algorithms, and Risk Aversion” within the priority pro- gramme SPP1962 “Non-smooth and Complementarity-based Distributed Parameter Systems: Simulation and Hierarchical Optimization”. Edited by Weierstraß-Institut für Angewandte Analysis und Stochastik (WIAS) Leibniz-Institut im Forschungsverbund Berlin e. V. Mohrenstraße 39 10117 Berlin Germany Fax: +49 30 20372-303 E-Mail: [email protected] World Wide Web: http://www.wias-berlin.de/ On the convexity of optimal control problems involving non-linear PDEs or VIs and applications to Nash games Michael Hintermüller, Steven-Marian Stengl Abstract Generalized Nash equilibrium problems in function spaces involving PDEs are considered. One of the central issues arising in this context is the question of existence, which requires the topological characterization of the set of minimizers for each player of the associated Nash game.
    [Show full text]
  • Ph. D. Programs for the Past Departments Or Schools Would Drop Colleagues Twenty Years and Are Not Typical
    ---------------------------------------------------------~hlIMPORTANT ADVISORY FOR GRADUATE STUDENTS IN MATHEMATICS In the past several years the jolJ prospects for ,1 recent recipient of a Ph D. degree in the mathematical sciences have changed from excellent to poor. All indications are that this rapid deterioration will accelerate over the next few years, and that the decade 1974-19H5 will be exceedingly bleak. The report of the chairman of the AMS Committee on Employment and Educational Policy should be read by everyone who is plfrnning to enter a career in mathematics in order to become acquainted with the facts concerning present and future employment. The first part of the report appeared in the November l!J7:l issue of these c\~/m•JJ, pages 34ti-352, and the second part appears in this issue, pages :lfi7-:l72. The action taken in response to these facts must, of course, be determined on an individual basis. Many will feel that for them mathematics is a calling and a way of life that must be followed at whatever cost and sacrifice. They will be undeterred by the prospect of difficulties and, hopefully, will be successful in finding suitable career opportunities. The Americ<m Mathematical Society, whose primary purpose is the fostering of mathematical research, will, of course, make every effort to ameliorate con­ ditions during the difficult period ahead. But it must be realized that the problem has its roots in complex social conditions which the Society has only the most limited powers to control or affect. Hichani S. Palais, Member AMS Committee on Employment and Educational Policv OF THE AMERICAN MATHEMATICAL SOCIETY Everett Pitcher and Gordon L.
    [Show full text]
  • On Norm Attaining Operators and Multilinear Maps
    On norm attaining operators and multilinear maps Francisco J. Falcó Benavent Supervised by Domingo García, Manuel Maestre & Pilar Rueda Facultat de Ciències Matemàtiques Universitat de València This dissertation is submitted for the degree of Doctor in Mathematics (University of Valencia) September 2014 I, Francisco J. Falcó Benavent, declare that this thesis titled, “On norm attaining operators and multilinear maps” and the work presented in it are my own. I confirm that: ■ This work was done wholly or mainly while in candidature for a degree of Doctor in Mathematics (University of Valencia) at this University. ■ Where I have consulted the published work of others, this is always clearly attributed. ■ Where I have quoted from the work of others, the source is always given. With the exception of such quotations, this thesis is entirely my own work. ■ I have acknowledged all main sources of help. Signed: Date: i To Maika for all the cups of coffee during the long nights. To Domingo, Manolo and Pilar for their dedication and wise advice. Acknowledgements “If I have seen further it is by standing on the shoulders of giants.” — Isaac Newton I would like to thank my supervisors for their support throughout these years, keeping me going when times were tough, asking insightful questions, and offering invaluable advice. They allowed me the space and freedom I needed to work and continued supporting and guiding me from day one. Their advice was essential to the completion of this dissertation and has taught me innumerable lessons and insights on the workings of academic research in general. I acknowledge with great gratitude Richard and Eleanor Aron, for so many things that if I had to write all them here, this thesis will be written in two volumes.
    [Show full text]
  • Juan Carlos Ferrando Manuel López-Pellicer Editors Descriptive Topology and Functional Analysis in Honour of Jerzy Kąkol’S 60Th Birthday
    16 Springer Proceedings in Mathematics & Statistics Justificación de ser editores. Referencias al Prometeo en referencias 17 y 20, capítulos 1 y 6 del libro. Juan Carlos Ferrando Manuel López-Pellicer Editors Descriptive Topology and Functional Analysis In Honour of Jerzy Kąkol’s 60th Birthday Datos del libro al dorso Editors Juan Carlos Ferrando Manuel López-Pellicer The University Research Institute Center Matematica Aplicada of Operations Research Universidad Politecnica de Valencia Universidad Miguel Hernandez Valencia Elche Spain Alicante Spain ISSN 2194-1009 ISSN 2194-1017 (electronic) ISBN 978-3-319-05223-6 ISBN 978-3-319-05224-3 (eBook) DOI 10.1007/978-3-319-05224-3 Springer Cham Heidelberg New York Dordrecht London Library of Congress Control Number: 2014942526 Ó Springer International Publishing Switzerland 2014 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer.
    [Show full text]
  • Boundaries for Spaces of Holomorphic Functions on C(K)
    Publ. RIMS, Kyoto Univ. 42 (2006), 27–44 Boundaries for Spaces of Holomorphic Functions on C(K) By Mar´ıa D. Acosta∗ Abstract We consider the Banach space Au(X) of holomorphic functions on the open unit ball of a (complex) Banach space X which are uniformly continuous on the closed unit ball, endowed with the supremum norm. A subset B of the unit ball of X is a boundary A ∈A | | for u(X) if for every F u(X), the norm of F is given by F =supx∈B F (x) . We prove that for every compact K, the subset of extreme points in the unit ball of C(K) is a boundary for Au(C(K)). If the covering dimension of K is at most one, then every norm attaining function in Au(C(K)) must attain its norm at an extreme point of the unit ball of C(K). We also show that for any infinite K, there is no Shilov boundary for Au(C(K)), that is, there is no minimal closed boundary, a result known before for K scattered. §1. Introduction A classical result by Silovˇ [Lo, p. 80] states that if A is a separating algebra of continuous functions on a compact Hausdorff space K, then there is a smallest closed subset F ⊂ K with the property that every function of A attains its maximum absolute value at some point of F . Bishop [Bi] proved that for every compact metrizable Hausdorff space K, any separating Banach algebra A ⊂C(K) has a minimal boundary, that is, there is M ⊂ K such that every element in A attains its norm at M and M is minimal with such a property.
    [Show full text]
  • Positively Convex Modules and Ordered Normed Linear Spaces
    Journal of Convex Analysis Volume 10 (2003), No. 1, 109–127 Positively Convex Modules and Ordered Normed Linear Spaces Dieter Pumpl¨un Fachbereich Mathematik, FernUniversit¨at,58084 Hagen, Germany [email protected] Dedicated to Harald Holmann on the occasion of his 70th birthday Received October 11, 2001 A positively convex module is a non-empty set closed under positively convex combinations but not necessarily a subset of a linear space. Positively convex modules are a natural generalization of positively convex subsets of linear spaces. Any positively convex module has a canonical semimetric and there is a universal positively aáne mapping into a regularly ordered normed linear space and a universal completion. 2000 Mathematics Subject Classification: 52A05, 52A01, 46B40, 46A55 1. Introduction In the following all linear spaces will be real. If E is a linear space ordered by a proper cone C; E = C C, and normed by a Riesz norm with respect to C, the positive part of the unit ball− O(E), i.e. O(E) C, is a typical example of a positively convex set or module. Actually any positively\ convex set or module i.e. any non-empty set P closed under positively convex operations, is an aáne quotient of such a positive part of a unit ball. Positively convex modules are natural generalizations of positively convex sets, the postulate that they are subsets of some linear space is dropped. Special positively convex sets play an important part in the theory of order-unit Banach spaces as so-called “universal capsÔ (cp. [21], §9) and also as generators of the topology of locally solid ordered topological linear spaces and the theory of Riesz seminorms (cp.
    [Show full text]
  • The Bishop-Phelps-Bollobás Theorem for L(Lp,Lq)
    THE BISHOP-PHELPS-BOLLOBAS´ THEOREM FOR OPERATORS ON L1(µ) YUN SUNG CHOI, SUN KWANG KIM, HAN JU LEE, AND MIGUEL MART´IN Dedicated to the memory of Robert R. Phelps Abstract. In this paper we show that the Bishop-Phelps-Bollob´astheorem holds for L(L1(µ);L1(ν)) for all measures µ and ν and also holds for L(L1(µ);L1(ν)) for every arbitrary measure µ and every localizable measure ν. Finally, we show that the Bishop-Phelps-Bollob´astheorem holds for two classes of bounded linear operators from a real L1(µ) into a real C(K) if µ is a finite measure and K is a compact Hausdorff space. In particular, one of the classes includes all Bochner representable operators and all weakly compact operators. 1. Introduction The celebrated Bishop-Phelps theorem of 1961 [12] states that for a Banach space X, every element in its dual space X∗ can be approximated by ones that attain their norms. Since then, there has been an extensive research to extend this result to bounded linear operators between Banach spaces [14, 27, 33, 34, 37] and non-linear mappings [2,7, 11, 16, 17, 30]. On the other hand, Bollob´as[13], motivated by problems arising in the theory of numerical ranges, sharpened the Bishop-Phelps theorem in 1970, and got what is nowadays called the Bishop-Phelps-Bollob´astheorem. Previously to presenting this result, let us introduce some notations. Given a (real or complex) Banach space X, we write BX for ∗ the unit ball, SX for its unit sphere, and X for the topological dual space of X.
    [Show full text]
  • A Revised Pre-Order Principle and Set-Valued Ekeland Variational
    A revised pre-order principle and set-valued Ekeland variational principle1 Jing-Hui Qiu School of Mathematical Sciences, Soochow University, Suzhou 215006, P. R. China Abstract. In my former paper “A pre-order principle and set-valued Ekeland variational principle” (see: arXiv: 1311.4951[math.FA]), we established a general pre-order principle. From the pre-order prin- ciple, we deduced most of the known set-valued Ekeland variational principles (denoted by EVPs) and their improvements. But the pre- order principle could not imply Khanh and Quy’s EVP in [On gener- alized Ekeland’s variational principle and equivalent formulations for set-valued mappings, J. Glob. Optim., 49 (2011), 381-396], where the perturbation contains a weak τ-function. In this paper, we give a re- vised version of the pre-order principle. This revised version not only implies the original pre-order principle, but also can be applied to obtain the above Khanh and Quy’s EVP. Thus, the revised pre-order principle implies all the known set-valued EVPs in set containing forms (to my knowledge). arXiv:1405.1522v1 [math.FA] 7 May 2014 Key words: Pre-order Principle, Ekeland variational principle, Set- valued map, Perturbation, Locally convex space, Vector optimization Mathematics Subject Classifications (2000) 49J53 · 90C48 · 65K10 · 46A03 1. Introduction In [20], we established a general pre-order principle, which consists of a pre- order set (X, ) and a monotone extended real-valued function η on (X, ). The pre-order principle states that there exists a strong minimal point dominated by any 1This work was supported by the National Natural Science Foundation of China (10871141).
    [Show full text]
  • Review in Studia Universitatis Babes–Bolyai, Series Mathematica, March
    Review in Studia Universitatis Babe»s{Bolyai, Series Mathematica, March 2003 C. Z¸alinescu, Convex Analysis in General Vector Spaces, World Scienti¯c, New Jersey{ London{Singapore{Hong Kong 2002, xx+367 pp, ISBN: 981-238-067-1 This well written book is devoted to convex analysis in in¯nite dimensional spaces. What makes it di®erent from other existing books on convex analysis and optimization is the fact that the results are presented in their most generality, known at this time, as well as the inclusion of new and recent material. The author is a well known specialist in the ¯eld and the book incorporates many of his original results. In order to obtain this generality, the author has included in the ¯rst chapter of the book, Ch. 1, Preliminary results on functional analysis, a detailed study of convex series (cs) closed, lower cs-closed (lcs-closed), ideally convex, lower ideally convex (li-convex) sets and multivalued mappings, allowing him to prove very general open mapping and closed graph theorems of Ursescu{ Robinson type. The chapter contains also a ¯ne study of separation of convex sets and a presentation, with complete proofs, of Ekeland's variational principle and of Borwein{ Preiss smooth variational principle. The second chapter of the book, Ch. 2, Convex analysis in locally convex spaces, beside classical results, contains also the study of some more general classes of functions, corre- sponding to the sets studied in the ¯rst chapter: cs-closed, cs-complete, lcs-closed, ideally convex, bcs-complete and li-convex functions. The conjugate function, duality formulae, the subdi®erential and the "-subdi®erential calculus, are also included.
    [Show full text]
  • Convex Functions: Constructions, Characterizations and Counterexamples C 20081
    Convex Functions: Constructions, Characterizations and Counterexamples c 20081 Jonathan M. Borwein and Jon Vanderwerff August 26, 2008 1Do not circulate without permission 2 To our wives. Judith and Judith. ii Preface This book on convex functions emerges out of fifteen years of collaboration be- tween the authors. It is by far from the first on the subject nor will it be the last. It is neither a book on Convex analysis such as Rockafellar’s foundational 1970 book [342] nor a book on Convex programming such as Boyd and Vanden- berghe’s excellent recent text [121]. There are a number of fine books—both recent and less so—on both those subjects or on Convexity and relatedly on Variational analysis. Books such as [344, 236, 351, 237, 115, 90, 295, 304] com- plement or overlap in various ways with our own focus which is to explore the interplay between the structure of a normed space and the properties of convex functions which can exist thereon. In some ways, among the most similar books to ours are those of Phelps [320] and of Giles [214] in that both also straddle the fields of geometric functional analysis and convex analysis—but without the convex function itself being the central character. We have structured this book so as to accommodate a variety of readers. This leads to some intentional repetition. Chapter one makes the case for the ubiquity of convexity, largely by way of example; many but not all of which are followed up in later chapters. Chapter two then provides a foundation for the study of convex functions in Euclidean (finite-dimensional) space, and Chap- ter three reprises important special structures such as polyhedrality, eigenvalue optimization and semidefinite programming.
    [Show full text]