• Documents
  • Authors
  • Tables
  • Log in
  • Sign up
  • MetaCart
  • DMCA
  • Donate

CiteSeerX logo

Advanced Search Include Citations
Advanced Search Include Citations

Global optimization of polynomials using gradient tentacles and sums of squares

by Markus Schweighofer
Venue:SIAM Journal on Optimization
Add To MetaCart

Tools

Sorted by:
Results 1 - 10 of 26
Next 10 →

Sparse SOS relaxations for minimizing functions that are summations of small polynomials

by Jiawang Nie, James Demmel - SIAM Journal On Optimization , 2008
"... This paper discusses how to find the global minimum of functions that are summations of small polynomials (“small ” means involving a small number of variables). Some sparse sum of squares (SOS) techniques are proposed. We compare their computational complexity and lower bounds with prior SOS relaxa ..."
Abstract - Cited by 23 (4 self) - Add to MetaCart
This paper discusses how to find the global minimum of functions that are summations of small polynomials (“small ” means involving a small number of variables). Some sparse sum of squares (SOS) techniques are proposed. We compare their computational complexity and lower bounds with prior SOS relaxations. Under certain conditions, we also discuss how to extract the global minimizers from these sparse relaxations. The proposed methods are especially useful in solving sparse polynomial system and nonlinear least squares problems. Numerical experiments are presented, which show that the proposed methods significantly improve the computational performance of prior methods for solving these problems. Lastly, we present applications of this sparsity technique in solving polynomial systems derived from nonlinear differential equations and sensor network localization. Key words: Polynomials, sum of squares (SOS), sparsity, nonlinear least squares, polynomial system, nonlinear differential equations, sensor network localization 1
(Show Context)

Citation Context

... Sturmfels [23] proposed to use SOS representations of f(x) − γ modulo the gradient ideal of f(x), and show that the minimum value f ∗ can be obtained when f ∗ is attained at some point. Schweighofer =-=[31]-=- proposes to minimize f(x) over a semialgebraic set called the gradient tentacle, and shows that the minimum value f ∗ can be computed when f ∗ > −∞ but not attainable. Jibetean and Laurent [13], and ...

Representation of non-negative polynomials, degree bounds and applications to optimization

by M. Marshall , 2006
"... Natural sufficient conditions for a polynomial to have a local minimum at a point are considered. These conditions tend to hold with probability 1. It is shown that polynomials satisfying these conditions at each minimum point have nice presentations in terms of sums of squares. Applications are giv ..."
Abstract - Cited by 22 (2 self) - Add to MetaCart
Natural sufficient conditions for a polynomial to have a local minimum at a point are considered. These conditions tend to hold with probability 1. It is shown that polynomials satisfying these conditions at each minimum point have nice presentations in terms of sums of squares. Applications are given to optimization on a compact set and also to global optimization. In many cases, there are degree bounds for such presentations. These bounds are of theoretical interest, but they appear to be too large to be of much practical use at present. In the final section, other more concrete degree bounds are obtained which ensure at least that the feasible set of solutions is not empty.
(Show Context)

Citation Context

...emark that the assumption that f achieves a minimum value on R n is restrictive. The minimum value of f on Z(I) need not equal the infimum of f on R n , e.g., consider f(x, y) = x 2 + (xy − 1) 2 . In =-=[23]-=- it is explained how 7the algorithm in [10] can be modified to handle the case where f is bounded from below on R n but does not achieve a minimum value on R n . 3. degree bounds in the compact case ...

An exact Jacobian SDP relaxation for polynomial optimization

by Jiawang Nie - Mathematical Programming, Series A
"... Given polynomials f(x), gi(x), hj(x), we study how to minimize f(x) on the set S = {x ∈ Rn: h1(x) = · · · = hm1(x) = 0, g1(x) ≥ 0,..., gm2(x) ≥ 0}. Let fmin be the minimum of f on S. Suppose S is nonsingular and fmin is achievable on S, which are true generically. This paper proposes a new t ..."
Abstract - Cited by 21 (7 self) - Add to MetaCart
Given polynomials f(x), gi(x), hj(x), we study how to minimize f(x) on the set S = {x ∈ Rn: h1(x) = · · · = hm1(x) = 0, g1(x) ≥ 0,..., gm2(x) ≥ 0}. Let fmin be the minimum of f on S. Suppose S is nonsingular and fmin is achievable on S, which are true generically. This paper proposes a new type semidefinite programming (SDP) relaxation which is the first one for solving this problem exactly. First, we con-struct new polynomials ϕ1,..., ϕr, by using the Jacobian of f, hi, gj, such that the above problem is equivalent to min x∈Rn f(x) s.t. hi(x) = 0, ϕj(x) = 0, 1 ≤ i ≤ m1, 1 ≤ j ≤ r, g1(x) ν1 · · · gm2(x)νm2 ≥ 0, ∀ν ∈ {0, 1}m2. Second, we prove that for all N big enough, the standard N-th order Lasserre’s SDP relaxation is exact for solving this equivalent problem, that is, its optimal value is equal to fmin. Some variations and examples are also shown.
(Show Context)

Citation Context

...rrect lower bound. Generally, this case will not happen, as shown by items (d), (e) of Theorem 2.5. In unconstrained optimization, when 23 fmin is not achievable, excellent approaches are proposed in =-=[9, 10, 26]-=-. It is an interesting future work to generalize them to constrained optimization. An important question is for what concrete relaxation order N∗ the SDP relaxation (2.8) is exact for solving (1.1). N...

An effective implementation of a symbolic-numeric cylindrical algebraic decomposition for quantifier elimination.

by Hidenao Iwane , Hitoshi Yanami , Hirokazu Anai - In Proceedings of the 2009 International Workshop on Symbolic-Numeric Computation, , 2009
"... ABSTRACT With many applications in engineering and in scientific fields, quantifier elimination (QE) has been attracting more attention these days. Cylindrical algebraic decomposition (CAD) is used as a basis for a general QE algorithm. We propose an effective symbolic-numeric cylindrical algebraic ..."
Abstract - Cited by 11 (0 self) - Add to MetaCart
ABSTRACT With many applications in engineering and in scientific fields, quantifier elimination (QE) has been attracting more attention these days. Cylindrical algebraic decomposition (CAD) is used as a basis for a general QE algorithm. We propose an effective symbolic-numeric cylindrical algebraic decomposition (SNCAD) algorithm for solving polynomial optimization problems. The main ideas are a bounded CAD construction approach and utilization of sign information. The bounded CAD constructs CAD only in restricted admissible regions to remove redundant projection factors and avoid lifting cells where truth values are constant over the region. By utilization of sign information we can avoid symbolic computation in the lifting phase. Techniques for implementation are also presented. These techniques help reduce the computing time. We have examined our implementation by solving many example problems. Experimental results show that our implementation significantly improves efficiency compared to our previous work.
(Show Context)

Citation Context

...truth values and utilization of quantifier information of input formulas. It significantly reduces the computing time in the lifting phase. Also, many approaches for special types of input formulas have been proposed, for example [30, 32, 33, 40, 5]. Meanwhile, several attempts to develop symbolic-numeric 168 algorithms for solving polynomial optimization problems (POPs) have been done. In [2] an efficient algorithm based on CAD to solve semidefinite programming exactly is studied. Another promising direction is to use generalized critical values for global optimization, which originates with [38] in the scheme of sum of squares. Then within the framework of the critical point method, generalized critical values are first used in [16]. As for sums of squares approach, we note that it provides algebraic certificates to compute lower bounds on global infimum (see [35]). In [26] computation of “symbolic” algebraic certificates (instead of “numerical” ones) is proposed. Moreover, the problem of obtaining theorems on the existence of algebraic certificates based on sums of squares is due to [20]. In this paper, we introduce an efficient approach for solving POPs by using SNCAD. Our approach...

Global Optimization of Polynomials Using Generalized Critical Values and Sums of Squares

by Feng Guo, Mohab Safey Ei Din, Lihong Zhi , 2010
"... Let ¯ X = [X1,..., Xn] and f ∈ R [ ¯ X]. We consider the problem of computing the global infimum of f when f is bounded below. For A ∈ GLn(C), we denote by f A the polynomial f(A ¯ X). Fix a number M ∈ R greater than infx∈Rn f(x). We prove that there exists a Zariski-closed subset A � GLn(C) such t ..."
Abstract - Cited by 11 (1 self) - Add to MetaCart
Let ¯ X = [X1,..., Xn] and f ∈ R [ ¯ X]. We consider the problem of computing the global infimum of f when f is bounded below. For A ∈ GLn(C), we denote by f A the polynomial f(A ¯ X). Fix a number M ∈ R greater than infx∈Rn f(x). We prove that there exists a Zariski-closed subset A � GLn(C) such that for all A ∈ GLn(Q) \ A, we have f A ≥ 0 on R n if and only if for all ɛ> 0, there exist sums of squares of polynomials s and t in R [ ¯ X] and polynomials φi ∈ R [ ¯ X] such that f A + ɛ = s + t ( M − f A) + ∑ A ∂f 1≤i≤n−1 φi. ∂Xi Hence we can formulate the original optimization problems as semidefinite programs which can be solved efficiently in Matlab. Some numerical experiments are given. We also discuss how to exploit the sparsity of SDP problems to overcome the ill-conditionedness of SDP problems when the infimum is not attained.

A new look at nonnegativity on closed sets and polynomial optimization

by Jean B. Lasserre - SIAM J. Optim , 2011
"... ar ..."
Abstract - Cited by 9 (4 self) - Add to MetaCart
Abstract not found
(Show Context)

Citation Context

...equence of upper bounds converges to f∗ even if f∗ is not attained, which when K = Rn, could provide an alternative and/or a complement to the hierarchy of convex relaxations provided in Schweighofer =-=[21]-=- (based on gradient tentacles) and in Hà and Pham [8] (based on the truncated tangency variety), which both provide again a monotone sequence of lower bounds. Finally, we also give a very simple inte...

Practical and theoretical issues for the computation of generalized critical values of a polynomial mapping

by Mohab Safey El Din
"... ..."
Abstract - Cited by 5 (1 self) - Add to MetaCart
Abstract not found

NEW APPROXIMATIONS FOR THE CONE OF COPOSITIVE MATRICES AND ITS DUAL

by Jean B. Lasserre
"... Abstract. We provide convergent hierarchies for the convex cone C of copositive matrices and its dual C ∗ , the cone of completely positive matrices. In both cases the corresponding hierarchy consists of nested spectrahedra and provide outer (resp. inner) approximations for C (resp. for its dual C ∗ ..."
Abstract - Cited by 4 (0 self) - Add to MetaCart
Abstract. We provide convergent hierarchies for the convex cone C of copositive matrices and its dual C ∗ , the cone of completely positive matrices. In both cases the corresponding hierarchy consists of nested spectrahedra and provide outer (resp. inner) approximations for C (resp. for its dual C ∗), thus complementing previous inner (resp. outer) approximations for C (for C ∗). In particular, both inner and outer approximations have a very simple interpretation. Finally, extension to K-copositivity and K-complete positivity for a closed convex cone K, is straightforward. hal-00545755, version 2- 19 Jan 2012 1.

CERTIFIED RELAXATION FOR POLYNOMIAL OPTIMIZATION ON SEMI-ALGEBRAIC SETS

by Marta Abril Bucero, Bernard Mourrain , 2013
"... In this paper, we describe a relaxation method to compute the minimal critical value of a real polynomial function on a semialgebraic set S and the ideal defining the points at which the minimal critical value is reached. We show that any relaxation hierarchy which is the projection of the Karush-K ..."
Abstract - Cited by 2 (0 self) - Add to MetaCart
In this paper, we describe a relaxation method to compute the minimal critical value of a real polynomial function on a semialgebraic set S and the ideal defining the points at which the minimal critical value is reached. We show that any relaxation hierarchy which is the projection of the Karush-Kuhn-Tucker relaxation stops in a finite number of steps and the ideal defining the minimizers is generated by the kernel of the associated moment matrix in that degree. Assuming the minimizer ideal is zerodimensional, we give a new criterion to detect when the minimum is reached and we prove that this criterion is satisfied for a sufficiently high degree. This exploits new representation of positive polynomials as elements of the preordering modulo the KKT ideal, which only involves polynomials in the initial set of variables.
(Show Context)

Citation Context

...fined by a radical ideal is also analyzed under similar regularity conditions. The case where the minimum value is not reached at points which satisfied the KKT constraints, has also been studied. In =-=[33]-=-, relaxation techniques are studied for functions for which the minimum is not reached and which satisfies some special properties “at infinity”. In [8], tangency constraints are used in a relaxation ...

PROBABILISTIC ALGORITHM FOR POLYNOMIAL OPTIMIZATION OVER A REAL ALGEBRAIC SET

by Aurélien Greuet, Mohab Safey El Din , 2013
"... Let f,f1,...,fs be polynomials with rational coefficients in the indeterminates X = X1,...,Xn of maximum degree D and V be the set of common complex solutions of F = (f1,...,fs). We give an algorithm which, up to some regularity assumptions on F, computes an exact representation of the global infi ..."
Abstract - Cited by 2 (2 self) - Add to MetaCart
Let f,f1,...,fs be polynomials with rational coefficients in the indeterminates X = X1,...,Xn of maximum degree D and V be the set of common complex solutions of F = (f1,...,fs). We give an algorithm which, up to some regularity assumptions on F, computes an exact representation of the global infimum f ⋆ = inf x∈V∩Rnf (x), i.e. a univariate polynomial vanishing at f ⋆ and an isolating interval for f ⋆. Furthermore, this algorithm decides whether f ⋆ is reached and if so, it returns x ⋆ ∈ V ∩Rn such that f (x⋆) = f ⋆. This algorithm is probabilistic. It makes use of the notion of polar varieties. Its complexity is essentially cubic in (sD) n and linear in the complexity of evaluating the input. This fits within the best known deterministic complexity class DO(n). We report on some practical experiments of a first implementation that is available as a Maple package. It appears that it can tackle global optimization problems that were unreachable by previous exact algorithms and can manage instances that are hard to solve with purely numeric techniques. As far as we know, even under the extra genericity assumptions on the input, it is the first probabilistic algorithm that combines practical efficiency with good control of complexity for this problem.
Powered by: Apache Solr
  • About CiteSeerX
  • Submit and Index Documents
  • Privacy Policy
  • Help
  • Data
  • Source
  • Contact Us

Developed at and hosted by The College of Information Sciences and Technology

© 2007-2019 The Pennsylvania State University