DOCSLIB.ORG

The Constant Trace Property in Noncommutative Optimization ISSAC’21, July 2021, Saint Petersburg, Russia

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
The Constant Trace Property in Noncommutative Optimization ISSAC’21, July 2021, Saint Petersburg, Russia The Constant Trace Property in Noncommutative Optimization Ngoc Hoang Anh Mai Abhishek Bhardwaj Victor Magron LAAS LAAS LAAS Toulouse, France Toulouse, France Toulouse, France [email protected] [email protected] [email protected] ABSTRACT A standard approach in the commutative setting, is Lasserre’s In this article, we show that each semidefinite relaxation of a Hierarchy [12], which provides a sequence lower bounds on the op- ball-constrained noncommutative polynomial optimization prob- timal values for POPs, with guaranteed convergence under some lem can be cast as a semidefinite program with a constant trace ma- natural constraints according to Putinar’s Positivstellensatz [24]. trix variable. We then demonstrate how this constant trace prop- This hierarchy and its nc extension to eigenvalue/trace optimiza- erty can be exploited via first order numerical methods to solve ef- tion [5, 23], involve solving SDPs over the space of multivariate ficiently the semidefinite relaxations of the noncommutative prob- moment and nc Hankel matrices, respectively. lem. Due to the current capacity of interior-point SDP solvers such as Mosek [1, 20], these hierarchies can only be applied when the mul- KEYWORDS tivariate moment (or nc Hankel) matrices are of “moderate” size. Often restricting their use to polynomials of low degrees, or in few noncommutative polynomial optimization, sums of hermitian squares, variables, with the situation being worse in the nc setting. eigenvalue and trace optimization, conditional gradient-based aug- A strategy for reducing the size of the SDP hierarchies is to mented Lagrangian, constant trace property, semidefinite program- exploit the sparsity structures of POPs. They include correlative ming sparsity (CS) in [11] and term sparsity (TS), CS-TS in [28] all of which are the analogs of the commutative works about CS [27], ACM Reference Format: Ngoc Hoang Anh Mai, Abhishek Bhardwaj, and Victor Magron. 2021. The TS [29, 30] and CS-TS [31]. Constant Trace Property in Noncommutative Optimization. In Proceedings Encouraged by [7, 33], in [17, 18] the first and third authors of ACM Conference (ISSAC’21). ACM, New York, NY, USA, 8 pages. showed how to exploit the Constant Trace Property (CTP) for SDP relaxations of POPs, which is satisfied when the matrices involved in the SDP relaxations have constant trace. By utilizing first or- 1 INTRODUCTION der spectral methods to solve the required SDP relaxations, they Polynomial optimization problems (POP) are present in many ar- attained significant computational gains for POPs constrained on eas of mathematics, and science in general. There are many appli- simple domains, e.g., sphere, ball, annulus, box and simplex. cations in global optimization, control and analysis of dynamical In this article, we extend the exploitation of the CTP to NCPOPs. systems to name a few [13], and being able to efficiently solve POP Our two main contributions are the following: First, we obtain anal- is of great importance. ogous results to [17, 18], which ensure the CTP for a broad class of In this article we focus on noncommutative (nc) polynomial opti- dense NCPOPs. In particular, if nc ball (or nc polydisc) constraint(s) mization problems (NCPOP), that is, polynomial optimization with is present, then CTP holds. We also extend this CTP-framework to non-commuting variables. NCPOP has several applications in con- some NCPOPs with correlative sparsity. Secondly, We provide a Ju- trol [26] and quantum information [6, 19, 21]. lia package for solving NCPOPs with CTP. The package makes use Since the advent of interior point methods for semidefinite pro- of first order methods for solving SDPs with CTP. We also demon- grams (SDP) [2], there have been many approaches to solving POP, strate the numerical and computational efficiency of this approach, using powerful representation results from real algebraic geometry on some sample classes of dense NCPOPs and NCPOPs with cor- arXiv:2102.02162v1 [math.OC] 3 Feb 2021 for positive polynomials. Inspired by Schmüdgen’s solution to the relative sparsity. moment problem on compact semialgebraic sets [25], these meth- ods aim to provide certificates of global positivity. There are natu- 2 DEFINITIONS & PRELIMINARIES ral analogues to these approaches in the nc setting, coming from Here we introduce some basic preliminary knowledge needed free algebraic geometry [9], and the tracial moment problems [4]. in the sequel. For a more detailed introduction to the topics intro- duced in this section, the reader is referred to [5]. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full cita- 2.1 Noncommutative polynomials tion on the first page. Copyrights for components of this work owned by others than We denote by - the noncommuting letters -1,...,-=. Let - = ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or re- h i -1,...,-= be the free monoid generated by -, and call its ele- publish, to post on servers or to redistribute to lists, requires prior specific permission h i and/or a fee. Request permissions from [email protected]. ments words in -. Given a word F = -81 ...-8A , F∗ is its reverse, ISSAC’21, July 2021, Saint Petersburg, Russia R i.e., F∗ = -8A ...-81 . Consider the free algebra - of polynomi- © 2021 Association for Computing Machinery. R h i ACM ISBN 978-x-xxxx-xxxx-x/YY/MM...$15.00 als in - with coefficients in . Its elements are called noncommu- https://doi.org/10.1145/nnnnnnn.nnnnnnn tative (nc) polynomials. Endow R - with the involution 5 5 h i → ∗ ISSAC’21, July 2021, Saint Petersburg, Russia Mai, Bhardwaj and Magron which fixes R - pointwise. The length of the longest word We say that & g h is Archimedean if for all @ R - , there is ∪ ( ) + ( ) ∈ h i in a polynomial 5 R - is called the degree of 5 and is de- a positive ' N such that ' @ @ & g h . ∈ h i ∈ − ∗ ∈ ( ) + ( ) noted deg 5 . We write R - for all nc polynomials of degree ( ) h i3 2.2.2 Semialgebraic sets. We define the semialgebraic set associ- at most 3. The set of symmetric elements of R - is defined as h i ated to g as SymR - = 5 R - : 5 ∗ = 5 . We employ the graded lexico- h i ∈ h i = graphic ordering on all structures and objects we consider. g = S : 6 g, 6 0 . D ∈ ∀ ∈ ( ) We write - for the set of all words in - of degree at most h i3 h i We can naturally extend this notion from matrix tuples of the same 3, and we let W - W be the column vector of words in - , 3 ( )≡ 3 h i3 order, to bounded self-adjoint operators on some Hilbert space , and V - V the column vector of words of degree 3. We also H 3 ( )≡ 3 which make 6 psd for all 6 g. This extension is called the denote by W3 (resp. V3 ) the set of all entries of W3 - (resp. ( ) ∈ ( ) operator semialgebraic set associated to g, and we denote it as g∞. V - ). The length of W is equal to s 3,= := 3 =8, which D 3 ( ) 3 ( ) 8=0 Similarly we define the variety associated to h as we write as s 3 , when contextually appropriate. Given a poly- = ( ) s 3 Í h = S : ℎ h,ℎ = 0 , nomial 5 R - 3 , let f = 5F F W R ( ) be its vector of ∈ h i ( ) ∈ 3 ∈ D ∈ ∀ ∈ ( ) coefficients. It is clear that every polynomial 5 R - 3 is of and the natural extension to the operator variety . ∈ h i h∞ the form 5 = 5 F = f W = W f. For 5 R - let D F W3 F ∗ 3 3∗ ∈ ∈ h i 2.2.3 Hankel matrices and the Riesz functional. Suppose we have 5 = deg 5 2 , and given some : N, we define :5 := : 5 , ⌈ ⌉ ⌈ ( )/Í⌉ ∈ < −⌈ ⌉ e.g., W = W . We use standard notations on R , i.e., given a truncated real valued sequence y = ~F F W23 . For each such : 5 :5 ( ) ∈ R< −⌈ ⌉ sequence, we define the Riesz functional, !y : R - R as a , a 2 denotes the usual 2-norm of a. h i23 → ∈ SAk k !y @ := @F~F for @ = @FF R - . Let denote the space of real symmetric matrices of size A, we ( ) F F ∈ h i23 Suppose further that y satisfies~F = ~F for allF W23 . We as- will normally omit the subscript A when we discuss matrices of Í Í ∗ ∈ S sociate to such y the nc Hankel matrix of order 3, M y , defined as arbitrary size, or if the size is clear from context. Given A , 3 ( ) ∈ M y D,E = !y D E , where D,E W . Given @ SymR - , we A is positive semidefinite (psd) (resp. positive definite (pd)), if all ( 3 ( )) ( ∗ ) ∈ 3 ∈ h i define the localizing matrix M @y as M @y D,E = !y D @E , eigenvalues of A are non-negative (resp. positive), and we write 3@ ( ) ( 3@ ( )) ( ∗ ) A where now D,E W . A 0 (resp. A 0). We denote by Tr A the trace ( 8,8) of 3@ ≻ ( ) 8=1 ∈ the matrix A SA and tr A = 1 Tr A is the normalized trace. Let ∈ ( ) A ( ) Í S (resp. S ) be the cone of psd (resp. pd) matrices. For a subset 2.3 Eigenvalue minimization + S, we++ define := S and := S . We write Given 5 SymR - , g, h SymR - , the minimal eigenvalue S ⊆ S+= S∩ + S++ S∩ ++ ∈ h i ⊂ h i = A1,..., A= S , and given @ R - , by @ we mean the of 5 over g∞ ∞ is given by: ( ) ∈ ∈ h i ( ) D ∩ Dh evaluation of @ - on , i.e., replacement of the nc letters -8 with ( ) the matrices A8 .
Load more

Recommended publications
  • Algebraic, Geometric, and Topological Methods in Optimization
    ALGEBRAIC,GEOMETRIC, AND TOPOLOGICAL METHODS IN OPTIMIZATION Jesús A. De Loera University of California, Davis 2019 — Joint Mathematical Meetings, Baltimore 1 This talk is about USING ALGEBRA, GEOMETRY and TOPOLOGY in in OPTIMIZATION research Due to time I will focus onLINEAR OPTIMIZATION... THE STORY OF THIS LECTURE -Optimization is a vibrant branch of Applied Mathematics. Its goal is to maximizing or minimizing some objective function relative to a set of possible solutions! E.g., Maximum profit, optimal arrangement, minimal error, etc. 2 - This talk is about USING ALGEBRA, GEOMETRY and TOPOLOGY in in OPTIMIZATION research Due to time I will focus onLINEAR OPTIMIZATION... THE STORY OF THIS LECTURE Optimization is a vibrant branch of Applied Mathematics. Its goal is to maximizing or minimizing some objective function relative to a set of possible solutions! E.g., Maximum profit, optimal arrangement, minimal error, etc. 2 - Due to time I will focus onLINEAR OPTIMIZATION... THE STORY OF THIS LECTURE Optimization is a vibrant branch of Applied Mathematics. Its goal is to maximizing or minimizing some objective function relative to a set of possible solutions! E.g., Maximum profit, optimal arrangement, minimal error, etc. This talk is about USING ALGEBRA, GEOMETRY and TOPOLOGY in in OPTIMIZATION research 2 - THE STORY OF THIS LECTURE Optimization is a vibrant branch of Applied Mathematics. Its goal is to maximizing or minimizing some objective function relative to a set of possible solutions! E.g., Maximum profit, optimal arrangement, minimal error, etc. This talk is about USING ALGEBRA, GEOMETRY and TOPOLOGY in in OPTIMIZATION research Due to time I will focus onLINEAR OPTIMIZATION..
    [Show full text]
  • Real Geometric Invariant Theory and Ricci Soliton Metrics on Two-Step Nilmanifolds
    Real Geometric Invariant Theory and Ricci Soliton Metrics on Two-step Nilmanifolds Michael R. Jablonski A dissertation submitted to the faculty of the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Mathematics. Chapel Hill 2008 Approved by Advisor: Patrick B. Eberlein Reader: James Damon Reader: Shrawan Kumar Reader: Michael Taylor Reader: Jonathan Wahl ABSTRACT MICHAEL R. JABLONSKI: Real Geometric Invariant Theory and Ricci Soliton Metrics on Two-step Nilmanifolds (Under the direction of Patrick B. Eberlein) In this work we study Real Geometric Invariant Theory and its applications to left-invariant geometry of nilpotent Lie groups. We develop some new results in the real category that distinguish GIT over the reals from GIT over the complexes. Moreover, we explore some of the basic relationships between real and complex GIT over projective space to obtain analogues of the well-known relationships that previously existed in the affine setting. This work is applied to the problem of finding left-invariant Ricci soliton metrics on two-step nilpotent Lie groups. Using our work on Real GIT, we show that most two-step nilpotent Lie groups admit left- invariant Ricci soliton metrics. Moreover, we build many new families of nilpotent Lie groups which cannot admit such metrics. ii To the Lord my God. iii ACKNOWLEDGEMENTS I would like to thank my thesis advisor, Dr. Patrick Eberlein, who introduced me to many interesting and worthwhile avenues of mathematics; his encouragement to not limit myself to one narrow field has made all the difference.
    [Show full text]
  • Quantitative Properties of Real Algebraic Singularities
    Ph. D thesis : Quantitative properties of real algebraic singularities Lionel F. Alberti Directeur de th`ese : Bernard Mourrain Codirecteur de th`ese : Georges Comte September 22, 2008 Contents 1 Introduction 4 1.1 Fran¸cais................................ 4 1.2 English ................................ 7 2 Fast and certified topology computations for planar curves 11 2.1 Notationsanddefinitions . 13 2.2 Overview ............................... 14 2.3 Regulardomains ........................... 15 2.4 Simplysingulardomains. 19 2.4.1 TopologicalDegree . 19 2.4.2 Counting the number of branches . 21 2.4.3 Conic structure and connection algorithm . 22 2.5 Isolatingtheinterestingpoints . 24 2.5.1 Subdivisionmethod . 25 2.5.2 Rational univariate representation . 27 2.6 Examples ............................... 28 3 Triangulating smooth algebraic varieties 31 3.1 Backgroundonmeshing implicit surfaces. 31 3.2 Algebraicingredients. 34 3.2.1 Representationofpolynomials . 34 3.2.2 Univariatesolver . 35 3.3 Towardaguaranteedmethod . 36 3.3.1 Descriptionofthealgorithm. 36 3.3.2 Complexityanalysis . 40 3.3.3 Singularities.......................... 42 3.4 Experimentation ........................... 43 4 Whitney stratifications, Transversality, and Triangulations 45 4.1 Stratification and topological stability . ... 51 4.1.1 Notations and background on stratifications . 51 4.1.2 An extension of the moving the wall theorem . 57 4.2 Measuringtransversality. 67 4.3 Conicstructure ............................ 72 1 4.3.1 The classical setup of the conic structure theorem . 72 4.3.2 Astableversionofconicstructure . 75 4.4 Relating Euclidean distance and transversality . .... 79 4.5 A tentative triangulation procedure . 84 5 A sweeping method to triangulate surfaces in three-space 93 5.1 Notations ............................... 94 5.1.1 Thecontourcurve .. .. .. .. .. .. .. .. .. 95 5.1.2 Characteristicpointsonthesurface . 96 5.2 Topologyofthex-criticalsections.
    [Show full text]
  • Real Algebraic Geometry Reading Group Topics MPI Leipzig July 3–7, 2017
    Real algebraic geometry reading group topics MPI Leipzig July 3{7, 2017. The following are proposed topics for talks to be delivered by the participants of the workshop. Further suggestions from you the participants are also welcome! Topology of real algebraic varieties • Constructing real algebraic varieties with interesting topology This talk should briefly describe the construction methods of Harnack, Hilbert, and Gudkov [Utk78] and an explanation of the classification of planar sextics. Then give a description of Viro's patchworking construction and theorem [Vir01] from the point of view of tropical geometry [IMS09]. This could also include Haas' patchworking theorem for curves [BIMS15, Chapter 3.3]. • Topology of real K3 surfaces This talk should describe Nikulin's classification of real K3 surfaces and planar sextics up to rigid isotopy via lattice theory [Nik80]. If time permits the speaker could also mention the classification of 1-nodal planar sextics by Itenberg by similar methods [Ite94]. • Topology of real algebraic space curves While the topology of planar curves has been extensively studied, the sudy of the topology of real algebraic curves in RP3 remained relatively unexamined. This talk should address the rigid isotopy classification of smooth irreducible algebraic curves in the real projective 3-space for small degree and genus [Bj11, MO16] and of canonically embedded curves of genus four [DZ99]. Convex algebraic geometry • Spectrahedral shadows Spectrahedral shadows are the feasible sets of semidefinite programming and thus the qustion of their characterisation is of high interest. A large class of semi-algebraic convex sets are in fact spectrahedral shadows as shown in [HN09, HN10].
    [Show full text]
  • Tame Topology and O-Minimal Structures, by Lou Van Den Dries, Cambridge Univ
    BULLETIN (New Series) OF THE AMERICAN MATHEMATICAL SOCIETY Volume 37, Number 3, Pages 351{357 S 0273-0979(00)00866-1 Article electronically published on March 2, 2000 Tame topology and o-minimal structures, by Lou van den Dries, Cambridge Univ. Press, New York, 1998, x + 180 pp., $39.95, ISBN 0-521-59838-9 x1 Semialgebraic and subanalytic geometry In classical algebraic geometry over an algebraically closed field, when an affine algebraic variety is projected to a lower dimensional affine space, the image is a finite boolean combination of algebraic varieties. The situation is more complicated when studying algebraic varieties over the real field. For example, when the circle x2 + y2 = 1 is projected on x-axis the image is the closed interval [−1; 1]. Thus when studying solutions to polynomial equations over the reals, we must also study polynomial inequalities. We say that a subset of Rn is semialgebraic if it is a finite Boolean combination of solution sets to polynomial equations p(X1;:::;Xn)=0 and polynomial inequalities q(X1;:::;Xn) > 0: It is easy to see that any semialgebraic set arises as the projection of an algebraic variety in a higher dimensional real affine space. One might worry that even more complicated sets could arise when we project semialgebraic sets, but this does not occur. Indeed, the Tarski-Seidenberg Principle guarantees that if X ⊆ Rn is semialgebraic and f : Rn ! Rm is a polynomial map, then the image of X under f is semialgebraic. One consequence of the Tarski-Seidenberg Principle is that the closure of a semialgebraic set is semialgebraic.
    [Show full text]
  • The Six Grothendieck Operations on O-Minimal Sheaves
    THE SIX GROTHENDIECK OPERATIONS ON O-MINIMAL SHEAVES MARIO´ J. EDMUNDO AND LUCA PRELLI Abstract. In this paper we develop the formalism of the Grothendieck six operations on o-minimal sheaves. The Grothendieck formalism allows us to obtain o-minimal versions of: (i) derived projection formula; (ii) universal coefficient formula; (iii) derived base change formula; (iv) K¨unnethformula; (v) local and global Verdier duality. 1. Introduction The study of o-minimal structures ([16]) is the analytic part of model theory which deals with theories of ordered, hence topological, structures satisfying cer- tain tameness properties. It generalizes piecewise linear geometry ([16, Chapter 1, x7]), semi-algebraic geometry ([4]) and globally sub-analytic geometry ([36], also called finitely sub-analytic in [15]) and it is claimed to be the formalization of Grothendieck's notion of tame topology (topologie mod´er´ee).See [16] and [18]. The most striking successes of this model-theoretic point of view of sub-analytic geometry include, on the one hand, an understanding of the behavior at infinity of certain important classes of sub-analytic sets as in Wilkie's ([55]) as pointed out by Bierstone and Milman [3], and on the other hand, the recent, somehow surprising, first unconditional proof of the Andr´e-Oortconjecture for mixed Shimura varieties expressible as products of curves by Pila [45] following previous work also using o-minimality by Pila and Zanier ([46]), Pila and Wilkie ([47]) and Peterzil and Starchenko ([43]). The goal of this paper is to contribute
    [Show full text]
  • (CO)HOMOLOGY 1. Introduction As It Is Known O-Minimal Geometry Is a Wide Ranging Gene
    TRANSACTIONS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 360, Number 9, September 2008, Pages 4889–4912 S 0002-9947(08)04403-6 Article electronically published on April 9, 2008 COMPARISON THEOREMS FOR O-MINIMAL SINGULAR (CO)HOMOLOGY MARIO´ J. EDMUNDO AND ARTHUR WOERHEIDE Abstract. Here we show the existence of the o-minimal simplicial and sin- gular (co)homology in o-minimal expansions of real closed fields and prove several comparison theorems for o-minimal (co)homology theories. 1. Introduction As it is known o-minimal geometry is a wide ranging generalization of semi- algebraic, sub-analytic and sub-pfaffian geometry. In this paper we will work over an o-minimal expansion N of an ordered ring (necessarily a real closed field) and definable means definable in N possibly with parameters. Associated to the first-order theory Th(N )ofN over N we have the categories of definable sets and of definably compact definable sets (or equivalently of closed and bounded definable sets; see [ps]). The morphisms on these categories are continuous definable maps. For basic o-minimality and the notion of definable sets and definable maps we refer the reader to [vdd]. By [KPS] and the first-order logic compactness theorem, for each infinite cardinal κ,thetheoryTh(N ) has a model of cardinality κ and any such model is also an o-minimal structure. Moreover, by [Sh], for every κ>max{ℵ0, |Th(N )|},there are up to isomorphism 2κ models M of Th(N ) such that |M| = κ.Amodel M of Th(N ) determines the categories of definable sets and of definably compact definable sets defined over M.
    [Show full text]
  • Real Algebraic Geometry Saugata Basu, Johannes Huisman, Kurdyka Krzysztof, Victoria Powers, Jean-Philippe Rolin
    Real Algebraic Geometry Saugata Basu, Johannes Huisman, Kurdyka Krzysztof, Victoria Powers, Jean-Philippe Rolin To cite this version: Saugata Basu, Johannes Huisman, Kurdyka Krzysztof, Victoria Powers, Jean-Philippe Rolin. Real Al- gebraic Geometry. Real Algberaic Geometry, Rennes, 2011, Jun 2011, Rennes, France. hal-00609687 HAL Id: hal-00609687 https://hal.archives-ouvertes.fr/hal-00609687 Submitted on 19 Jul 2011 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Conference Real Algebraic Geometry Université de Rennes 1 20 - 24 june 2011 Beaulieu Campus Conference Real Algebraic Geometry Université de Rennes 1 20 - 24 june 2011 Beaulieu Campus Conference Real Algebraic Geometry C’est un très grand plaisir pour nous d’accueillir à Rennes cette conférence internationale de Géométrie Algébrique Réelle qui est le quatrième opus d’une belle série commencée il y a 30 ans. La toute première conférence internationale de la discipline fut organisée à Rennes en 1981 par Jean-Louis Colliot-Thélène, Michel Coste, Louis Mahé et Marie-Françoise Roy. Depuis, la discipline a pris un essor considérable : de très nombreuses conférences sur la thématique se sont déroulées dans le monde entier, un réseau européen a tissé des liens au cœur duquel Rennes 1 a toujours joué un rôle privilégié.
    [Show full text]
  • An Introduction to Semialgebraic Geometry
    AN INTRODUCTION TO SEMIALGEBRAIC GEOMETRY Michel COSTE Institut de Recherche Math´ematiquede Rennes [email protected] Octobre 2002 Contents Introduction 3 1 Counting the real roots 5 1.1 Sturm’s theorem . ....................... 5 1.1.1 P without multiple root . ............... 5 1.1.2 P with multiple roots . ................... 6 1.1.3 A bound for the roots ................... 7 1.2 Real roots satisfying inequalities . ............... 7 1.2.1 One inequality ....................... 7 1.2.2 Several inequalities . ................... 9 1.2.3 Deciding the existence of a solution of a system of poly- nomial equations and inequalities . .......... 10 1.3 Systems with parameters . ................... 10 1.3.1 Tarski-Seidenberg . ................... 11 1.3.2 Systems with one equation . ............... 12 1.3.3 Example: polynomial of degree 4 . .......... 13 1.3.4 General systems ....................... 16 1.4 Another method for counting real roots .............. 17 1.4.1 Hermite’s method . ................... 17 1.4.2 Determination of the signature .............. 19 1.4.3 Principal subresultant coefficients . .......... 21 2 Semialgebraic sets 25 2.1 Stability properties . ....................... 25 2.1.1 Definition and first examples . ............... 25 2.1.2 Consequences of Tarski-Seidenberg principle . ...... 26 2.2 Semialgebraic functions ....................... 29 2.2.1 Definition and first properties ............... 29 2.2.2 TheLojasiewicz 8 inequality . ............... 30 1 2 CONTENTS 2.3 Decomposition of a semialgebraic set . .............. 32 2.3.1 Cylindrical algebraic decomposition . ......... 32 2.3.2 Construction of an adapted c.a.d. ............. 34 2.3.3 The c.a.d. algorithm . .................. 38 2.3.4 Connected components of semialgebraic sets . ..... 40 3 Triangulation of semialgebraic sets 41 3.1 Thom’s lemma .
    [Show full text]
  • Algebraic Geometry Minus Fields
    Algebraic geometry minus fields Thomas Scanlon University of California, Berkeley 20 March 2009 MSRI Thomas Scanlon University of California, Berkeley Algebraic geometry minus fields What is model theory? 1.1: What is model theory? Model theory is the branch of mathematical logic which deals with the relation between a for- mal language and its interpre- tations, or models. ... The line between universal algebra and model theory is sometimes fuzzy; our own usage is ex- plained by the equation universal algebra + logic = model theory. Thomas Scanlon University of California, Berkeley Algebraic geometry minus fields And what is model theory? And what is model theory? Model theory is about the classification of mathematical structures, maps and sets by means of logical formulas. One can classify structures accord- ing to what logical sentences are true in them; in fact the term ‘model’ comes from the usage ‘structure A is a model of sentence φ,’ meaning that φ is true in A. Thomas Scanlon University of California, Berkeley Algebraic geometry minus fields And what is model theory? In 1973 C. C. Chang and Jerry Keisler characterized model theory as universal algebra plus logic. They meant the universal algebra to stand for structures and the logic to stand for logical formulas. This is neat, but it might suggest that model theorists and universal algebraists have closely related interests, which is debatable. Also it leaves out the fact that model theorists study the sets definable in a single structure by a logical formula. In this respect model theorists are much closer to algebraic geometers, who study the sets of points definable by equations over a field.
    [Show full text]
  • Metric Algebraic Geometry
    Metric Algebraic Geometry Maddie Weinstein University of California Berkeley [email protected] 1 / 24 What is metric algebraic geometry? This is the title of my PhD dissertation. Metric algebraic geometry is concerned with properties of real algebraic varieties that depend on a distance metric. The distance metric can be the Euclidean metric in the ambient space or a metric intrinsic to the variety. Results can be applied in the computational study of the geometry of data with nonlinear models. 2 / 24 Application: Computational Study of Geometry of Data with Nonlinear Models n Suppose we are given of sample of points in R and we choose to model them with an algebraic variety V . What can we learn about the variety? In Learning Algebraic Varieties from Samples, joint with Paul Breiding, Sara Kalisnik, Bernd Sturmfels, we survey, develop, and implement methods to study the following properties: What is the dimension of V ? What equations vanish on V ? What is the degree of V ? What are the irreducible components of V ? What are the homology groups of V ? 3 / 24 Topological Data Analysis Figure: Persistent homology barcodes for the quartic Trott curve in R2. See Topology and Data by Gunnar Carlsson for more information. 4 / 24 Algebraicity of Persistent Homology In Offset Hypersurfaces and Persistent Homology of Algebraic Varieties, joint with Emil Horobet, we study the persistent homology of the offset filtration of an algebraic variety, bringing the perspective of real algebraic geometry to the study of persistent homology and showing that barcodes can be computed exactly. Figure: The quartic Viviani curve (left) and its = 1 offset surface (right), which is of degree 10.
    [Show full text]
  • Combinatorial Complexity in O-Minimal Geometry
    Submitted exclusively to the London Mathematical Society doi:10.1112/0000/000000 Combinatorial complexity in o-minimal geometry Saugata Basu Abstract In this paper we prove tight bounds on the combinatorial and topological complexity of sets defined in terms of n definable sets belonging to some fixed definable family of sets in an o- minimal structure. This generalizes the combinatorial parts of similar bounds known in the case of semi-algebraic and semi-Pfaffian sets, and as a result vastly increases the applicability of results on combinatorial and topological complexity of arrangements studied in discrete and computational geometry. As a sample application, we extend a Ramsey-type theorem due to Alon et al. [3], originally proved for semi-algebraic sets of fixed description complexity to this more general setting. 1. Introduction Over the last twenty years there has been a lot of work on bounding the topological complexity (measured in terms of their Betti numbers) of several different classes of subsets of Rk – most notably semi-algebraic and semi-Pfaffian sets. The usual setting for proving these bounds is as follows. One considers a semi-algebraic (or semi-Pfaffian) set S ⊂ Rk defined by a Boolean formula whose atoms consists of P > 0,P = 0, P < 0,P ∈ P, where P is a set of polynomials (resp. Pfaffian functions) of degrees bounded by a parameter (resp. whose Pfaffian complexity is bounded by certain parameters) and #P = n. It is possible to obtain bounds on the Betti numbers of S in terms of n, k and the parameters bounding the complexity of the functions in P.
    [Show full text]