Semidefinite Optimization using MOSEK
|
|
- Brent McDowell
- 5 years ago
- Views:
Transcription
1 Semidefinite Optimization using MOSEK Joachim Dahl ISMP Berlin, August 23,
2 Introduction Introduction Semidefinite optimization Algorithms Results and examples Summary MOSEK is a state-of-the-art solver for large-scale linear and conic quadratic problems. Based on the homogeneous model using Nesterov-Todd scaling. Next version includes semidefinite optimization. Goalforfirst versionis to beatsedumi. WhySDP inmosek? It is very powerful and flexible for modeling. We wantthesdp woryou developed to beavailableto our customers! Customers have been asing about it for awhile. 2 /28
3 Semidefinite optimization 3 /28
4 Semidefinite optimization Primal and dual problems: minimize p j=1 C j X j subjectto p j=1 A ij X j = b i X j S n j +, maximize b T y subjectto m i=1 y ia ij +S j = C j S j S n j +, wherea ij,c j S n j. Optimality conditions: p A ij X j = b i, j=1 m i=1 y i A ij +S j = C j, X j S j = 0, X j,s j S n j +. EquivalentcomplementaryconditionsifX j,s j S n j + : X j S j = 0 X j S j = 0 X j S j +S j X j = 0. 4 /28
5 Algorithms 5 /28
6 Notation For symmetricmatricesu S n we define svec(u) = (U 11, 2U 21,..., 2U n1,u 22, 2U 32,..., 2U n2,...,u nn ) T with inverse operation smat(u) = u 1 u 2 / 2 u n / 2 u 2 / 2 u n+1 u 2n 1 / 2... u n / 2 u n 1 / 2 u n(n+1)/2, and a symmetric product u v := (1/2)(smat(u)smat(v) + smat(v)smat(u)). 6 /28
7 Notation Let a ij = svec(a ij ), c j := svec(c j ), x j := svec(x j ), s j := svec(s j ), and A := a T a T 1p., c := c 1., x := x 1., s := s 1.. a T m1... a T mp c p x p s p We then have primal and dual problems: minimize c T x subjectto Ax = b x 0, maximize b T y subjectto A T y +s = c s 0. 7 /28
8 The homogeneous model Simplified homogeneous model: s 0 κ + 0 A T c A 0 b c T b T 0 x y τ = x,s 0, τ,κ 0. Note that (x,y,s,κ,τ) = 0isfeasible,and x T s+κτ = 0. Ifτ > 0, κ = 0 then(x,y,s)/τ is aprimal-dualoptimal solution, Ax = bτ, A T y +s = cτ, x T s = 0. Ifκ > 0, τ = 0 theneitherprimalor dualisinfeasible, Ax = 0, A T y +s = 0, c T x b T y < 0, Primalinfeasibleifb T y > 0, dualinfeasibleifc T x < 0. 8 /28
9 The central path Introduction Semidefinite optimization Algorithms Notation Homogeneous model The central path Primal-dual scaling The search direction Schur complement Practical details Results and examples Summary We definethe centralpath as: s 0 κ + 0 A T c A 0 b c T b T 0 x y τ = γ x s = γµ (0) e, τκ = γµ (0), whereγ [0,1],µ (0) = (x(0) ) T s (0) +κ (0) τ (0) n+1 > 0and e = svec(i n1 ). svec(i np ). A T y (0) +s (0) cτ (0) Ax (0) bτ (0) c T x (0) b T y (0) +κ (0) 9 /28
10 Primal-dual scaling Introduction Semidefinite optimization Algorithms Notation Homogeneous model The central path Primal-dual scaling The search direction Schur complement Practical details A primal-dualscalingisdefined byanon-singularr as W (x ) := svec(r T X R 1 ), W T (s ) := svec(r S R T ), preserves the semidefinite cone, x 0, s 0 W (x ) 0, W T (s ) 0, Results and examples Summary the central path, x s = γµe W (x ) W T (s ) = γµe, andthe complementarity: x T s = W (x ) T W T (s ). 10 /28
11 Two popular primal-dual scalings Introduction Semidefinite optimization Algorithms Notation Homogeneous model The central path Primal-dual scaling The search direction Schur complement Practical details Results and examples Summary Helmberg-Kojima-Monteiro(HKM)R 1 = S 1/2 : W (x ) = svec(s 1/2 X S 1/2 ), W T (s ) = e. Nesterov-Todd (NT): R 1 = ( ) 1/2, S 1/2 (S 1/2 X S 1/2 ) 1/2 S 1/2 Satisfies R 1 X R 1 = R S R. Computed as: R 1 = Λ 1/4 Q T L 1, LL T = X, QΛQ T = L T S L whichsatisfies R T X R 1 = R S R T = Λ. 11 /28
12 The search direction Introduction Semidefinite optimization Algorithms Notation Homogeneous model The central path Primal-dual scaling The search direction Schur complement Practical details Results and examples Summary Linearizing the scaled centrality condition W (x ) W T (s ) = γµe and discarding higher-order terms leads to where x +Π s = γµs 1 x, HKM scaling: Π z = svec(x Z S 1 +S 1 Z X )/2, NT scaling: Π z = svec(r T R Z R T R ). 12 /28
13 The search direction Most expensive part of interior-point optimizer is solving: A x b τ = (γ 1)(Ax bτ) = r 1 A T y + s c τ = (γ 1)(A T y +s cτ) = r 2 c T x b T y + κ = (γ 1)(c T x b T y +κ) = r 3 x+π s = γµs 1 x = r 4 κ+κ/τ τ = γµ/τ κ = r 5 Gives constant decrease in residuals and complementarity: A(x+α x) b(τ +α τ) = (1 α(1 γ))(ax bτ) (x+α x) T (s+α s)+(κ+α κ)(τ +α τ) = (1 α(1 γ))(x T s+κτ). }{{} <1 Polynomial complexity for γ chosen properly. 13 /28
14 Solving for the search direction After bloc-elimination,weget a2 2system: AΠA T y (AΠc+b) τ = r y (AΠc b) y +(c T Πc+κ/τ) τ = r τ. We factoraπa T = LL T and eliminate y = L T L 1 ((AΠc+b) τ +r y ). Then ( (AΠc b) T L T L 1 (AΠc+b)+(c T Πc+κ/τ) ) τ = r τ (AΠc b) T L T L 1 r y. An extra(insignificant) solve compared to an infeasible method. 14 /28
15 Forming the Schur complement Schur complement computed as a sum of outer products: AΠA T = = a T a T 1p Π a T m1... a T mp p ( P A :, Π A T ) :, PT P, =1 P T Π p a a m1.. a 1p... a mp wherep is apermutationof A :,. Complexity depends on choiceofp. We computeonlytril(aπa T ), soby nnz(p A :, ) 1 nnz(p A :, ) 2 nnz(p A :, ) p, get a good reduction in complexity. More general than SDPA. 15 /28
16 Forming the Schur complement To evaluate each term a T i Π a j = (R T A ir ) (R T A jr ) = A i (R R T A jr R T ). we follow SeDuMi. Rewrite R R T A jr R T = ˆMM +M T ˆMT, i.e., a symmetric(low-ran) product. A i sparse: evaluate ˆMM +M T ˆMT elementbyelement. A i dense: evaluate ˆMM +M T ˆMT usinglevel3blas. May exploit low-ran structure later. 16 /28
17 Practical details Introduction Semidefinite optimization Algorithms Notation Homogeneous model The central path Primal-dual scaling The search direction Schur complement Practical details Results and examples Summary Additionally, the MOSEK solver uses Mehrotra s predictor-corrector step. solves mixed linear, conic quadratic and semidefinite problems. employs a presolve step for linear variables, which can reduce problem size and complexity significantly. scales constraints trying to improving conditioning of a problem. 17 /28
18 Results and examples 18 /28
19 SDPLIB benchmar Introduction Semidefinite optimization Algorithms Results and examples SDPLIB benchmar Profiling Conic modeling Summary We usethe subsetofsdplib usedbyh. D.Mittelmann for his SDP solver comparison. For brevity we report only iteration count, computation time, and relative gap at termination. Our SDPA-format converter detects bloc-diagonal structure within the semidefinite blocs(sedumi and SDPA does not), so forafewproblemsoursolution time is much smaller. All solversrunonasingle threadon anintel Xeon E32270with4coresat3.4GHz and32gbmemoryusing Linux. 19 /28
20 SDPLIB benchmar MOSEK 7.0 SeDuMi 1.3 SDPA Problem it time relgap it time relgap it time relgap arch e e e-09 control e e e-07 control e e e-06 control e e e-06 equalg e e e-07 equalg e e e-08 gpp e e e-09 gpp e e e-09 hinf e e e-02 maxg e e e-08 maxg e e e-08 maxg e e e-08 mcp e e e-08 mcp e e e-09 qap e e e-05 qap e e e-04 qpg e e e-08 qpg e e e-08 ss e e e-08 theta e e e-09 theta e e e-08 theta e e e-08 theta e e e-08 thetag e e e-08 thetag e e e-07 truss e e e-08 truss e e e-09 relgap = (c T x b T y)/(1+ c T x + b T y ) 20 /28
21 Profiling Profiling results [%] Problem Schur factor stepsize search dir update arch control control control equalg equalg gpp gpp hinf maxg maxg maxg mcp mcp qap qap qpg qpg ss theta theta theta theta thetag thetag truss truss /28
22 Profiling conclusions Introduction Semidefinite optimization Algorithms Results and examples SDPLIB benchmar Profiling Conic modeling Summary FormingAΠA T (Schur)isnot alwaysmost expensive step (as perhaps believed). A good parallelization speedup is possible for computingaπa T. (Factoris alreadyparallelized). The update step (updating variables, neighborhood chec,newnt scaling) isvery cheapforlpand SOCP, butexpensiveforsdp, mainlydueto NT scaling;hkm scaling would give an improvement. Stepsize computations(e.g., stepsize to boundary) are moderately expensive, but hard to speed up. 22 /28
23 Conic modeling Introduction Semidefinite optimization Algorithms Results and examples SDPLIB benchmar Profiling Conic modeling Summary For A S n, thenearest correlationmatrixis X = arg min X S+ n,diag(x)=e A X F. Conic formulation exploiting symmetry of A X: minimize t subjectto z 2 t svec(a X) = z diag(x) = e X 0. Simple formulation on paper, but complicated to formulate in standard form as in SeDuMi, SDPA, /28
24 Conic modeling Introduction Semidefinite optimization Algorithms Results and examples SDPLIB benchmar Profiling Conic modeling Summary For conic modeling you need a modeling tool! Nice MATLAB toolboxes exist (Yalmip, CVX,...), but not for other languages. We developed a modeling API called MOSEK Fusion: A tool forself-dualconicmodeling;no QPor general convex optimization. Idea: create and manipulate linear expressions, and assign them to different cones. Simple but flexible design; scales well for large problems. Available for Python, Java,.NET. Planned versions include MATLAB (soon), C++ (later). Syntax almost identical across platforms. 24 /28
25 Python/Fusion listing for nearest correlation def svec(e): N = e.get_shape().dim(0) S = Matrix.sparse(N * (N+1) / 2, N * N, range(n * (N+1) / 2), [ (i+j*n) for j in xrange(n) for i in xrange(j,n) ], [ (1.0 if i == j else 2**(0.5)) for j in xrange(n) for i in xrange(j,n) ]) return Expr.mul(S,Expr.reshape( e, N * N )) def nearestcorr(a): M = Model("NearestCorrelation") N = len(a) # Setting up the variables X = M.variable("X",Domain.inPSDCone(N)) # t > z _2 tz = M.variable("tz", Domain.inQCone(N*(N+1)/2+1)) t = tz.index(0) z = tz.slice(1,n*(n+1)/2+1) # svec (A-X) = z M.constraint( Expr.sub(svec(Expr.sub(DenseMatrix(A),X)), z), Domain.equalsTo(0.0) ) # diag(x) = e for i in range(n): M.constraint( X.index(i,i), Domain.equalsTo(1.0) ) # Objective: Minimize t M.objective(ObjectiveSense.Minimize, t) M.solve() 25 /28
26 Summary 26 /28
27 Summary Introduction Semidefinite optimization Algorithms Results and examples Summary References SDP solver in MOSEK 7.0 based on self-dual homogeneous embedding and NT scaling. Faster than SeDuMi, comparable with SDPA. Includes Fusion, a powerful tool for self-dual conic modeling. Future directions: Incorporate multithreading. Exploit low-ran structure in data. Possibly implement the HKM direction. Wantto tryit? Contact support@mose.com for a beta version. 27 /28
28 References [1] E.D. Andersen, C. Roos, and T. Terlay. On implementing a primal-dual interior-point method for conic quadratic optimization. Mathematical Programming, 95(2): , [2] Y.E. Nesterov and M.J. Todd. Primal-dual interior-point methods for self-scaled cones. SIAM Journal on Optimization, 8(2): , [3] J.F. Sturm. Implementation of interior point methods for mixed semidefinite and second order cone optimization problems. Optimization Methods and Software, 17(6): , [4] M. Yamashita, K. Fujisawa, and M. Kojima. Implementation and evaluation of sdpa 6.0(semidefinite programming algorithm 6.0). Optimization Methods and Software, 18(4): , [5] Y.Ye,M.J.Todd,and S.Mizuno. AnO( nl)-iterationhomogeneousand self-dual linear programming algorithm. Mathematics of Operations Research, pages 53 67, /28
Solving conic optimization problems with MOSEK
Solving conic optimization problems with MOSEK Erling D. Andersen MOSEK ApS, Fruebjergvej 3, Box 16, 2100 Copenhagen, Denmark. WWW: http://www.mosek.com August 8, 2013 http://www.mosek.com 2 / 42 Outline
More informationOn implementing a primal-dual interior-point method for conic quadratic optimization
On implementing a primal-dual interior-point method for conic quadratic optimization E. D. Andersen, C. Roos, and T. Terlaky December 18, 2000 Abstract Conic quadratic optimization is the problem of minimizing
More informationLecture 17: Primal-dual interior-point methods part II
10-725/36-725: Convex Optimization Spring 2015 Lecture 17: Primal-dual interior-point methods part II Lecturer: Javier Pena Scribes: Pinchao Zhang, Wei Ma Note: LaTeX template courtesy of UC Berkeley EECS
More informationImplementation and Evaluation of SDPA 6.0 (SemiDefinite Programming Algorithm 6.0)
Research Reports on Mathematical and Computing Sciences Series B : Operations Research Department of Mathematical and Computing Sciences Tokyo Institute of Technology 2-12-1 Oh-Okayama, Meguro-ku, Tokyo
More information18. Primal-dual interior-point methods
L. Vandenberghe EE236C (Spring 213-14) 18. Primal-dual interior-point methods primal-dual central path equations infeasible primal-dual method primal-dual method for self-dual embedding 18-1 Symmetric
More informationOn the Behavior of the Homogeneous Self-Dual Model for Conic Convex Optimization
Mathematical Programming manuscript No. 2016 Springer International Publishing AG. Part of Springer Nature. http://dx.doi.org/10.1007/s10107-005-0667-3 On the Behavior of the Homogeneous Self-Dual Model
More informationConvex Optimization : Conic Versus Functional Form
Convex Optimization : Conic Versus Functional Form Erling D. Andersen MOSEK ApS, Fruebjergvej 3, Box 16, DK 2100 Copenhagen, Blog: http://erlingdandersen.blogspot.com Linkedin: http://dk.linkedin.com/in/edandersen
More informationParallel implementation of primal-dual interior-point methods for semidefinite programs
Parallel implementation of primal-dual interior-point methods for semidefinite programs Masakazu Kojima, Kazuhide Nakata Katsuki Fujisawa and Makoto Yamashita 3rd Annual McMaster Optimization Conference:
More information15. Conic optimization
L. Vandenberghe EE236C (Spring 216) 15. Conic optimization conic linear program examples modeling duality 15-1 Generalized (conic) inequalities Conic inequality: a constraint x K where K is a convex cone
More informationThe Ongoing Development of CSDP
The Ongoing Development of CSDP Brian Borchers Department of Mathematics New Mexico Tech Socorro, NM 87801 borchers@nmt.edu Joseph Young Department of Mathematics New Mexico Tech (Now at Rice University)
More informationSPARSE SECOND ORDER CONE PROGRAMMING FORMULATIONS FOR CONVEX OPTIMIZATION PROBLEMS
Journal of the Operations Research Society of Japan 2008, Vol. 51, No. 3, 241-264 SPARSE SECOND ORDER CONE PROGRAMMING FORMULATIONS FOR CONVEX OPTIMIZATION PROBLEMS Kazuhiro Kobayashi Sunyoung Kim Masakazu
More informationPENNON A Code for Convex Nonlinear and Semidefinite Programming
PENNON A Code for Convex Nonlinear and Semidefinite Programming Michal Kočvara Michael Stingl Abstract We introduce a computer program PENNON for the solution of problems of convex Nonlinear and Semidefinite
More informationLecture 5. Theorems of Alternatives and Self-Dual Embedding
IE 8534 1 Lecture 5. Theorems of Alternatives and Self-Dual Embedding IE 8534 2 A system of linear equations may not have a solution. It is well known that either Ax = c has a solution, or A T y = 0, c
More informationPrimal-Dual Geometry of Level Sets and their Explanatory Value of the Practical Performance of Interior-Point Methods for Conic Optimization
Primal-Dual Geometry of Level Sets and their Explanatory Value of the Practical Performance of Interior-Point Methods for Conic Optimization Robert M. Freund M.I.T. June, 2010 from papers in SIOPT, Mathematics
More informationIntroduction to Semidefinite Programs
Introduction to Semidefinite Programs Masakazu Kojima, Tokyo Institute of Technology Semidefinite Programming and Its Application January, 2006 Institute for Mathematical Sciences National University of
More informationResearch overview. Seminar September 4, Lehigh University Department of Industrial & Systems Engineering. Research overview.
Research overview Lehigh University Department of Industrial & Systems Engineering COR@L Seminar September 4, 2008 1 Duality without regularity condition Duality in non-exact arithmetic 2 interior point
More informationA NEW SECOND-ORDER CONE PROGRAMMING RELAXATION FOR MAX-CUT PROBLEMS
Journal of the Operations Research Society of Japan 2003, Vol. 46, No. 2, 164-177 2003 The Operations Research Society of Japan A NEW SECOND-ORDER CONE PROGRAMMING RELAXATION FOR MAX-CUT PROBLEMS Masakazu
More informationInterior-Point Methods
Interior-Point Methods Stephen Wright University of Wisconsin-Madison Simons, Berkeley, August, 2017 Wright (UW-Madison) Interior-Point Methods August 2017 1 / 48 Outline Introduction: Problems and Fundamentals
More informationInfeasible Primal-Dual (Path-Following) Interior-Point Methods for Semidefinite Programming*
Infeasible Primal-Dual (Path-Following) Interior-Point Methods for Semidefinite Programming* Yin Zhang Dept of CAAM, Rice University Outline (1) Introduction (2) Formulation & a complexity theorem (3)
More informationIMPLEMENTATION OF INTERIOR POINT METHODS
IMPLEMENTATION OF INTERIOR POINT METHODS IMPLEMENTATION OF INTERIOR POINT METHODS FOR SECOND ORDER CONIC OPTIMIZATION By Bixiang Wang, Ph.D. A Thesis Submitted to the School of Graduate Studies in Partial
More information1. Introduction. We consider the (primal) semidefinite program in block diagonal
RIGOROUS ERROR BOUNDS FOR THE OPTIMAL VALUE IN SEMIDEFINITE PROGRAMMING CHRISTIAN JANSSON, DENIS CHAYKIN, AND CHRISTIAN KEIL Abstract. A wide variety of problems in global optimization, combinatorial optimization
More information1 Outline Part I: Linear Programming (LP) Interior-Point Approach 1. Simplex Approach Comparison Part II: Semidenite Programming (SDP) Concludin
Sensitivity Analysis in LP and SDP Using Interior-Point Methods E. Alper Yldrm School of Operations Research and Industrial Engineering Cornell University Ithaca, NY joint with Michael J. Todd INFORMS
More informationInfeasible Primal-Dual (Path-Following) Interior-Point Methods for Semidefinite Programming*
Infeasible Primal-Dual (Path-Following) Interior-Point Methods for Semidefinite Programming* Notes for CAAM 564, Spring 2012 Dept of CAAM, Rice University Outline (1) Introduction (2) Formulation & a complexity
More informationPENNON A Generalized Augmented Lagrangian Method for Convex NLP and SDP p.1/39
PENNON A Generalized Augmented Lagrangian Method for Convex NLP and SDP Michal Kočvara Institute of Information Theory and Automation Academy of Sciences of the Czech Republic and Czech Technical University
More informationA Generalized Homogeneous and Self-Dual Algorithm. for Linear Programming. February 1994 (revised December 1994)
A Generalized Homogeneous and Self-Dual Algorithm for Linear Programming Xiaojie Xu Yinyu Ye y February 994 (revised December 994) Abstract: A generalized homogeneous and self-dual (HSD) infeasible-interior-point
More informationNew stopping criteria for detecting infeasibility in conic optimization
Optimization Letters manuscript No. (will be inserted by the editor) New stopping criteria for detecting infeasibility in conic optimization Imre Pólik Tamás Terlaky Received: March 21, 2008/ Accepted:
More informationConic Linear Optimization and its Dual. yyye
Conic Linear Optimization and Appl. MS&E314 Lecture Note #04 1 Conic Linear Optimization and its Dual Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A.
More informationSecond-order cone programming
Outline Second-order cone programming, PhD Lehigh University Department of Industrial and Systems Engineering February 10, 2009 Outline 1 Basic properties Spectral decomposition The cone of squares The
More informationOn Two Measures of Problem Instance Complexity and their Correlation with the Performance of SeDuMi on Second-Order Cone Problems
2016 Springer International Publishing AG. Part of Springer Nature. http://dx.doi.org/10.1007/s10589-005-3911-0 On Two Measures of Problem Instance Complexity and their Correlation with the Performance
More informationAgenda. Interior Point Methods. 1 Barrier functions. 2 Analytic center. 3 Central path. 4 Barrier method. 5 Primal-dual path following algorithms
Agenda Interior Point Methods 1 Barrier functions 2 Analytic center 3 Central path 4 Barrier method 5 Primal-dual path following algorithms 6 Nesterov Todd scaling 7 Complexity analysis Interior point
More informationIII. Applications in convex optimization
III. Applications in convex optimization nonsymmetric interior-point methods partial separability and decomposition partial separability first order methods interior-point methods Conic linear optimization
More informationDistributed Control of Connected Vehicles and Fast ADMM for Sparse SDPs
Distributed Control of Connected Vehicles and Fast ADMM for Sparse SDPs Yang Zheng Department of Engineering Science, University of Oxford Homepage: http://users.ox.ac.uk/~ball4503/index.html Talk at Cranfield
More informationarxiv: v1 [math.oc] 26 Sep 2015
arxiv:1509.08021v1 [math.oc] 26 Sep 2015 Degeneracy in Maximal Clique Decomposition for Semidefinite Programs Arvind U. Raghunathan and Andrew V. Knyazev Mitsubishi Electric Research Laboratories 201 Broadway,
More informationLecture 5. The Dual Cone and Dual Problem
IE 8534 1 Lecture 5. The Dual Cone and Dual Problem IE 8534 2 For a convex cone K, its dual cone is defined as K = {y x, y 0, x K}. The inner-product can be replaced by x T y if the coordinates of the
More informationResearch Note. A New Infeasible Interior-Point Algorithm with Full Nesterov-Todd Step for Semi-Definite Optimization
Iranian Journal of Operations Research Vol. 4, No. 1, 2013, pp. 88-107 Research Note A New Infeasible Interior-Point Algorithm with Full Nesterov-Todd Step for Semi-Definite Optimization B. Kheirfam We
More informationInterior Point Methods. We ll discuss linear programming first, followed by three nonlinear problems. Algorithms for Linear Programming Problems
AMSC 607 / CMSC 764 Advanced Numerical Optimization Fall 2008 UNIT 3: Constrained Optimization PART 4: Introduction to Interior Point Methods Dianne P. O Leary c 2008 Interior Point Methods We ll discuss
More informationSDPARA : SemiDefinite Programming Algorithm PARAllel version
Research Reports on Mathematical and Computing Sciences Series B : Operations Research Department of Mathematical and Computing Sciences Tokyo Institute of Technology 2-12-1 Oh-Okayama, Meguro-ku, Tokyo
More informationHow to generate weakly infeasible semidefinite programs via Lasserre s relaxations for polynomial optimization
CS-11-01 How to generate weakly infeasible semidefinite programs via Lasserre s relaxations for polynomial optimization Hayato Waki Department of Computer Science, The University of Electro-Communications
More informationThe Q Method for Symmetric Cone Programmin
The Q Method for Symmetric Cone Programming The Q Method for Symmetric Cone Programmin Farid Alizadeh and Yu Xia alizadeh@rutcor.rutgers.edu, xiay@optlab.mcma Large Scale Nonlinear and Semidefinite Progra
More informationOn the solution of large-scale SDP problems by the modified barrier method using iterative solvers
Mathematical Programming manuscript No. (will be inserted by the editor) Michal Kočvara Michael Stingl On the solution of large-scale SDP problems by the modified barrier method using iterative solvers
More informationOn the solution of large-scale SDP problems by the modified barrier method using iterative solvers
Mathematical Programming manuscript No. (will be inserted by the editor) Michal Kočvara Michael Stingl On the solution of large-scale SDP problems by the modified barrier method using iterative solvers
More informationA full-newton step feasible interior-point algorithm for P (κ)-lcp based on a new search direction
Croatian Operational Research Review 77 CRORR 706), 77 90 A full-newton step feasible interior-point algorithm for P κ)-lcp based on a new search direction Behrouz Kheirfam, and Masoumeh Haghighi Department
More informationFast ADMM for Sum of Squares Programs Using Partial Orthogonality
Fast ADMM for Sum of Squares Programs Using Partial Orthogonality Antonis Papachristodoulou Department of Engineering Science University of Oxford www.eng.ox.ac.uk/control/sysos antonis@eng.ox.ac.uk with
More informationA priori bounds on the condition numbers in interior-point methods
A priori bounds on the condition numbers in interior-point methods Florian Jarre, Mathematisches Institut, Heinrich-Heine Universität Düsseldorf, Germany. Abstract Interior-point methods are known to be
More informationSEMIDEFINITE PROGRAM BASICS. Contents
SEMIDEFINITE PROGRAM BASICS BRIAN AXELROD Abstract. A introduction to the basics of Semidefinite programs. Contents 1. Definitions and Preliminaries 1 1.1. Linear Algebra 1 1.2. Convex Analysis (on R n
More informationA CONIC DANTZIG-WOLFE DECOMPOSITION APPROACH FOR LARGE SCALE SEMIDEFINITE PROGRAMMING
A CONIC DANTZIG-WOLFE DECOMPOSITION APPROACH FOR LARGE SCALE SEMIDEFINITE PROGRAMMING Kartik Krishnan Advanced Optimization Laboratory McMaster University Joint work with Gema Plaza Martinez and Tamás
More informationCroatian Operational Research Review (CRORR), Vol. 3, 2012
126 127 128 129 130 131 132 133 REFERENCES [BM03] S. Burer and R.D.C. Monteiro (2003), A nonlinear programming algorithm for solving semidefinite programs via low-rank factorization, Mathematical Programming
More informationCuts for Conic Mixed-Integer Programming
Cuts for Conic Mixed-Integer Programming Alper Atamtürk and Vishnu Narayanan Department of Industrial Engineering and Operations Research, University of California, Berkeley, CA 94720-1777 USA atamturk@berkeley.edu,
More informationWarmstarting the Homogeneous and Self-Dual Interior Point Method for Linear and Conic Quadratic Problems
Mathematical Programming Computation manuscript No. (will be inserted by the editor) Warmstarting the Homogeneous and Self-Dual Interior Point Method for Linear and Conic Quadratic Problems Anders Skajaa
More informationPrimal-Dual Interior-Point Methods. Javier Peña Convex Optimization /36-725
Primal-Dual Interior-Point Methods Javier Peña Convex Optimization 10-725/36-725 Last time: duality revisited Consider the problem min x subject to f(x) Ax = b h(x) 0 Lagrangian L(x, u, v) = f(x) + u T
More informationA PREDICTOR-CORRECTOR PATH-FOLLOWING ALGORITHM FOR SYMMETRIC OPTIMIZATION BASED ON DARVAY'S TECHNIQUE
Yugoslav Journal of Operations Research 24 (2014) Number 1, 35-51 DOI: 10.2298/YJOR120904016K A PREDICTOR-CORRECTOR PATH-FOLLOWING ALGORITHM FOR SYMMETRIC OPTIMIZATION BASED ON DARVAY'S TECHNIQUE BEHROUZ
More informationContinuous Optimisation, Chpt 9: Semidefinite Problems
Continuous Optimisation, Chpt 9: Semidefinite Problems Peter J.C. Dickinson DMMP, University of Twente p.j.c.dickinson@utwente.nl http://dickinson.website/teaching/2016co.html version: 21/11/16 Monday
More informationResearch Reports on Mathematical and Computing Sciences
ISSN 1342-284 Research Reports on Mathematical and Computing Sciences Exploiting Sparsity in Linear and Nonlinear Matrix Inequalities via Positive Semidefinite Matrix Completion Sunyoung Kim, Masakazu
More informationAdvances in Convex Optimization: Theory, Algorithms, and Applications
Advances in Convex Optimization: Theory, Algorithms, and Applications Stephen Boyd Electrical Engineering Department Stanford University (joint work with Lieven Vandenberghe, UCLA) ISIT 02 ISIT 02 Lausanne
More informationLinear Algebra and its Applications
Linear Algebra and its Applications 433 (2010) 1101 1109 Contents lists available at ScienceDirect Linear Algebra and its Applications journal homepage: www.elsevier.com/locate/laa Minimal condition number
More informationLecture 6: Conic Optimization September 8
IE 598: Big Data Optimization Fall 2016 Lecture 6: Conic Optimization September 8 Lecturer: Niao He Scriber: Juan Xu Overview In this lecture, we finish up our previous discussion on optimality conditions
More informationResearch Reports on Mathematical and Computing Sciences
ISSN 1342-2804 Research Reports on Mathematical and Computing Sciences Correlative sparsity in primal-dual interior-point methods for LP, SDP and SOCP Kazuhiro Kobayashi, Sunyoung Kim and Masakazu Kojima
More informationPrimal-dual IPM with Asymmetric Barrier
Primal-dual IPM with Asymmetric Barrier Yurii Nesterov, CORE/INMA (UCL) September 29, 2008 (IFOR, ETHZ) Yu. Nesterov Primal-dual IPM with Asymmetric Barrier 1/28 Outline 1 Symmetric and asymmetric barriers
More informationSMCP Documentation. Release Martin S. Andersen and Lieven Vandenberghe
SMCP Documentation Release 0.4.5 Martin S. Andersen and Lieven Vandenberghe May 18, 2018 Contents 1 Current release 3 2 Future releases 5 3 Availability 7 4 Authors 9 5 Feedback and bug reports 11 5.1
More informationSemidefinite Programming, Combinatorial Optimization and Real Algebraic Geometry
Semidefinite Programming, Combinatorial Optimization and Real Algebraic Geometry assoc. prof., Ph.D. 1 1 UNM - Faculty of information studies Edinburgh, 16. September 2014 Outline Introduction Definition
More informationAn Infeasible Interior-Point Algorithm with full-newton Step for Linear Optimization
An Infeasible Interior-Point Algorithm with full-newton Step for Linear Optimization H. Mansouri M. Zangiabadi Y. Bai C. Roos Department of Mathematical Science, Shahrekord University, P.O. Box 115, Shahrekord,
More informationLecture: Introduction to LP, SDP and SOCP
Lecture: Introduction to LP, SDP and SOCP Zaiwen Wen Beijing International Center For Mathematical Research Peking University http://bicmr.pku.edu.cn/~wenzw/bigdata2015.html wenzw@pku.edu.cn Acknowledgement:
More information4. Algebra and Duality
4-1 Algebra and Duality P. Parrilo and S. Lall, CDC 2003 2003.12.07.01 4. Algebra and Duality Example: non-convex polynomial optimization Weak duality and duality gap The dual is not intrinsic The cone
More informationFast Algorithms for SDPs derived from the Kalman-Yakubovich-Popov Lemma
Fast Algorithms for SDPs derived from the Kalman-Yakubovich-Popov Lemma Venkataramanan (Ragu) Balakrishnan School of ECE, Purdue University 8 September 2003 European Union RTN Summer School on Multi-Agent
More informationPOLYNOMIAL OPTIMIZATION WITH SUMS-OF-SQUARES INTERPOLANTS
POLYNOMIAL OPTIMIZATION WITH SUMS-OF-SQUARES INTERPOLANTS Sercan Yıldız syildiz@samsi.info in collaboration with Dávid Papp (NCSU) OPT Transition Workshop May 02, 2017 OUTLINE Polynomial optimization and
More informationA CONIC INTERIOR POINT DECOMPOSITION APPROACH FOR LARGE SCALE SEMIDEFINITE PROGRAMMING
A CONIC INTERIOR POINT DECOMPOSITION APPROACH FOR LARGE SCALE SEMIDEFINITE PROGRAMMING Kartik Krishnan Sivaramakrishnan Department of Mathematics NC State University kksivara@ncsu.edu http://www4.ncsu.edu/
More informationInterior Point Methods in Mathematical Programming
Interior Point Methods in Mathematical Programming Clóvis C. Gonzaga Federal University of Santa Catarina, Brazil Journées en l honneur de Pierre Huard Paris, novembre 2008 01 00 11 00 000 000 000 000
More informationInexact primal-dual path-following algorithms for a special class of convex quadratic SDP and related problems
Inexact primal-dual path-following algorithms for a special class of convex quadratic SDP and related problems K.C. Toh, R. H. Tütüncü, and M. J. Todd February 8, 005 Dedicated to the memory of Jos Sturm:
More informationOn the implementation and usage of SDPT3 a Matlab software package for semidefinite-quadratic-linear programming, version 4.0
On the implementation and usage of SDPT3 a Matlab software package for semidefinite-quadratic-linear programming, version 4.0 Kim-Chuan Toh, Michael J. Todd, and Reha H. Tütüncü Draft, 16 June 2010 Abstract
More informationא K ٢٠٠٢ א א א א א א E٤
المراجع References المراجع العربية K ١٩٩٠ א א א א א E١ K ١٩٩٨ א א א E٢ א א א א א E٣ א K ٢٠٠٢ א א א א א א E٤ K ٢٠٠١ K ١٩٨٠ א א א א א E٥ المراجع الا جنبية [AW] [AF] [Alh] [Ali1] [Ali2] S. Al-Homidan and
More informationExploiting Sparsity in Primal-Dual Interior-Point Methods for Semidefinite Programming.
Research Reports on Mathematical and Computing Sciences Series B : Operations Research Department of Mathematical and Computing Sciences Tokyo Institute of Technology 2-12-1 Oh-Okayama, Meguro-ku, Tokyo
More informationTamás Terlaky George N. and Soteria Kledaras 87 Endowed Chair Professor. Chair, Department of Industrial and Systems Engineering Lehigh University
5th SJOM Bejing, 2011 Cone Linear Optimization (CLO) From LO, SOCO and SDO Towards Mixed-Integer CLO Tamás Terlaky George N. and Soteria Kledaras 87 Endowed Chair Professor. Chair, Department of Industrial
More informationIdentifying Redundant Linear Constraints in Systems of Linear Matrix. Inequality Constraints. Shafiu Jibrin
Identifying Redundant Linear Constraints in Systems of Linear Matrix Inequality Constraints Shafiu Jibrin (shafiu.jibrin@nau.edu) Department of Mathematics and Statistics Northern Arizona University, Flagstaff
More informationA Homogeneous Interior-Point Algorithm for Nonsymmetric Convex Conic Optimization
Mathematical Programming manuscript No. (will be inserted by the editor) A Homogeneous Interior-Point Algorithm for Nonsymmetric Convex Conic Optimization Anders Skajaa Yinyu Ye Received: date / Accepted:
More informationA PARALLEL INTERIOR POINT DECOMPOSITION ALGORITHM FOR BLOCK-ANGULAR SEMIDEFINITE PROGRAMS IN POLYNOMIAL OPTIMIZATION
A PARALLEL INTERIOR POINT DECOMPOSITION ALGORITHM FOR BLOCK-ANGULAR SEMIDEFINITE PROGRAMS IN POLYNOMIAL OPTIMIZATION Kartik K. Sivaramakrishnan Department of Mathematics North Carolina State University
More informationLMI solvers. Some notes on standard. Laboratoire d Analyse et d Architecture des Systèmes. Denis Arzelier - Dimitri Peaucelle - Didier Henrion
Some notes on standard LMI solvers Denis Arzelier - Dimitri Peaucelle - Didier Henrion Laboratoire d Analyse et d Architecture des Systèmes LAAS-CNRS 5 L = h l 1 l i Motivation State-feedback stabilization:
More informationSparse Matrix Theory and Semidefinite Optimization
Sparse Matrix Theory and Semidefinite Optimization Lieven Vandenberghe Department of Electrical Engineering University of California, Los Angeles Joint work with Martin S. Andersen and Yifan Sun Third
More informationm i=1 c ix i i=1 F ix i F 0, X O.
What is SDP? for a beginner of SDP Copyright C 2005 SDPA Project 1 Introduction This note is a short course for SemiDefinite Programming s SDP beginners. SDP has various applications in, for example, control
More informationA PARALLEL INTERIOR POINT DECOMPOSITION ALGORITHM FOR BLOCK-ANGULAR SEMIDEFINITE PROGRAMS
A PARALLEL INTERIOR POINT DECOMPOSITION ALGORITHM FOR BLOCK-ANGULAR SEMIDEFINITE PROGRAMS Kartik K. Sivaramakrishnan Department of Mathematics North Carolina State University kksivara@ncsu.edu http://www4.ncsu.edu/
More informationSemidefinite Programming
Semidefinite Programming Basics and SOS Fernando Mário de Oliveira Filho Campos do Jordão, 2 November 23 Available at: www.ime.usp.br/~fmario under talks Conic programming V is a real vector space h, i
More informationSolving conic optimization problems via self-dual embedding and facial reduction: a unified approach
Solving conic optimization problems via self-dual embedding and facial reduction: a unified approach Frank Permenter Henrik A. Friberg Erling D. Andersen August 18, 216 Abstract We establish connections
More informationSolving large Semidefinite Programs - Part 1 and 2
Solving large Semidefinite Programs - Part 1 and 2 Franz Rendl http://www.math.uni-klu.ac.at Alpen-Adria-Universität Klagenfurt Austria F. Rendl, Singapore workshop 2006 p.1/34 Overview Limits of Interior
More informationLecture 1. 1 Conic programming. MA 796S: Convex Optimization and Interior Point Methods October 8, Consider the conic program. min.
MA 796S: Convex Optimization and Interior Point Methods October 8, 2007 Lecture 1 Lecturer: Kartik Sivaramakrishnan Scribe: Kartik Sivaramakrishnan 1 Conic programming Consider the conic program min s.t.
More informationInexact primal-dual path-following algorithms for a special class of convex quadratic SDP and related problems
Inexact primal-dual path-following algorithms for a special class of convex quadratic SDP and related problems K. C. Toh, R. H. Tütüncü, and M. J. Todd May 26, 2006 Dedicated to Masakazu Kojima on the
More informationTamás Terlaky George N. and Soteria Kledaras 87 Endowed Chair Professor. Chair, Department of Industrial and Systems Engineering Lehigh University
BME - 2011 Cone Linear Optimization (CLO) From LO, SOCO and SDO Towards Mixed-Integer CLO Tamás Terlaky George N. and Soteria Kledaras 87 Endowed Chair Professor. Chair, Department of Industrial and Systems
More informationContinuous Optimisation, Chpt 9: Semidefinite Optimisation
Continuous Optimisation, Chpt 9: Semidefinite Optimisation Peter J.C. Dickinson DMMP, University of Twente p.j.c.dickinson@utwente.nl http://dickinson.website/teaching/2017co.html version: 28/11/17 Monday
More informationSolving second order cone programming via a reduced augmented system approach
Solving second order cone programming via a reduced augmented system approach Zhi Cai and Kim-Chuan Toh Abstract The standard Schur complement equation based implementation of interior-point methods for
More informationA semidefinite relaxation scheme for quadratically constrained quadratic problems with an additional linear constraint
Iranian Journal of Operations Research Vol. 2, No. 2, 20, pp. 29-34 A semidefinite relaxation scheme for quadratically constrained quadratic problems with an additional linear constraint M. Salahi Semidefinite
More informationInterior Point Methods for Mathematical Programming
Interior Point Methods for Mathematical Programming Clóvis C. Gonzaga Federal University of Santa Catarina, Florianópolis, Brazil EURO - 2013 Roma Our heroes Cauchy Newton Lagrange Early results Unconstrained
More informationA Full Newton Step Infeasible Interior Point Algorithm for Linear Optimization
A Full Newton Step Infeasible Interior Point Algorithm for Linear Optimization Kees Roos e-mail: C.Roos@tudelft.nl URL: http://www.isa.ewi.tudelft.nl/ roos 37th Annual Iranian Mathematics Conference Tabriz,
More informationOn the solution of large-scale SDP problems by the modified barrier method using iterative solvers
Mathematical Programming manuscript No. (will be inserted by the editor) Michal Kočvara Michael Stingl On the solution of large-scale SDP problems by the modified barrier method using iterative solvers
More informationA direct formulation for sparse PCA using semidefinite programming
A direct formulation for sparse PCA using semidefinite programming A. d Aspremont, L. El Ghaoui, M. Jordan, G. Lanckriet ORFE, Princeton University & EECS, U.C. Berkeley A. d Aspremont, INFORMS, Denver,
More informationConvex Optimization. (EE227A: UC Berkeley) Lecture 6. Suvrit Sra. (Conic optimization) 07 Feb, 2013
Convex Optimization (EE227A: UC Berkeley) Lecture 6 (Conic optimization) 07 Feb, 2013 Suvrit Sra Organizational Info Quiz coming up on 19th Feb. Project teams by 19th Feb Good if you can mix your research
More informationA full-newton step infeasible interior-point algorithm for linear programming based on a kernel function
A full-newton step infeasible interior-point algorithm for linear programming based on a kernel function Zhongyi Liu, Wenyu Sun Abstract This paper proposes an infeasible interior-point algorithm with
More informationSecond Order Cone Programming Relaxation of Nonconvex Quadratic Optimization Problems
Second Order Cone Programming Relaxation of Nonconvex Quadratic Optimization Problems Sunyoung Kim Department of Mathematics, Ewha Women s University 11-1 Dahyun-dong, Sudaemoon-gu, Seoul 120-750 Korea
More informationEnlarging neighborhoods of interior-point algorithms for linear programming via least values of proximity measure functions
Enlarging neighborhoods of interior-point algorithms for linear programming via least values of proximity measure functions Y B Zhao Abstract It is well known that a wide-neighborhood interior-point algorithm
More informationLecture 10. Primal-Dual Interior Point Method for LP
IE 8534 1 Lecture 10. Primal-Dual Interior Point Method for LP IE 8534 2 Consider a linear program (P ) minimize c T x subject to Ax = b x 0 and its dual (D) maximize b T y subject to A T y + s = c s 0.
More informationThe Q Method for Second-Order Cone Programming
The Q Method for Second-Order Cone Programming Yu Xia Farid Alizadeh July 5, 005 Key words. Second-order cone programming, infeasible interior point method, the Q method Abstract We develop the Q method
More informationImplementation of Interior Point Methods for Mixed Semidefinite and Second Order Cone Optimization Problems
Implementation of Interior Point Methods for Mixed Semidefinite and Second Order Cone Optimization Problems Jos F. Sturm Department of Econometrics, Tilburg University, The Netherlands. j.f.sturm@kub.nl
More informationApproximate Farkas Lemmas in Convex Optimization
Approximate Farkas Lemmas in Convex Optimization Imre McMaster University Advanced Optimization Lab AdvOL Graduate Student Seminar October 25, 2004 1 Exact Farkas Lemma Motivation 2 3 Future plans The
More information