Directed subdifferentiable functions and the directed subdifferential without Delta-convex structure
- Authors: Baier, Robert , Farkhi, Elza , Roshchina, Vera
- Date: 2014
- Type: Text , Journal article
- Relation: Journal of Optimization Theory and Applications Vol. 160, no. 2 (2014), p. 391-414
- Full Text: false
- Reviewed:
- Description: We show that the directed subdifferential introduced for differences of convex (delta-convex, DC) functions by Baier and Farkhi can be constructed from the directional derivative without using any information on the delta-convex structure of the function. The new definition extends to a more general class of functions, which includes Lipschitz functions definable on o-minimal structure and quasidifferentiable functions. © 2013 Springer Science+Business Media New York.
Facially exposed cones are not always nice
- Authors: Roshchina, Vera
- Date: 2014
- Type: Text , Journal article
- Relation: SIAM Journal on Optimization Vol. 24, no. 1 (2014), p. 257-268
- Full Text:
- Reviewed:
- Description: We address the conjecture proposed by Gábor Pataki that every facially exposed cone is nice. We show that the conjecture is true in the three-dimensional case; however, there exists a four-dimensional counterexample of a cone that is facially exposed but is not nice.
Fast computation of zeros of polynomial systems with bounded degree under finite-precision
- Authors: Briquel, Irenee , Cucker, Felipe , Peña, Javier , Roshchina, Vera
- Date: 2014
- Type: Text , Journal article
- Relation: Mathematics of Computation Vol. 83, no. 287 (2014), p. 1279-1317
- Full Text: false
- Reviewed:
- Description: A solution for Smale's 17th problem, for the case of systems with bounded degree was recently given. This solution, an algorithm computing approximate zeros of complex polynomial systems in average polynomial time, assumed infinite precision. In this paper we describe a finite-precision version of this algorithm. Our main result shows that this version works within the same time bounds and requires a precision which, on the average, amounts to a polynomial amount of bits in the mantissa of the intervening floating-point numbers. © 2013 American Mathematical Society.
From the Farkas lemma to the Hahn-Banach theorem
- Authors: Dinh, Nguyen , Goberna, Miguel , López, Marco , Mo, T. H.
- Date: 2014
- Type: Text , Journal article
- Relation: SIAM Journal on Optimization Vol. 24, no. 2 (2014), p. 678-701
- Full Text:
- Reviewed:
- Description: This paper provides new versions of the Farkas lemma characterizing those inequalities of the form f(x) ≥ 0 which are consequences of a composite convex inequality (S ° g)(x) ≤ 0 on a closed convex subset of a given locally convex topological vector space X, where f is a proper lower semicontinuous convex function defined on X, S is an extended sublinear function, and g is a vector-valued S-convex function. In parallel, associated versions of a stable Farkas lemma, considering arbitrary linear perturbations of f, are also given. These new versions of the Farkas lemma, and their corresponding stable forms, are established under the weakest constraint qualification conditions (the so-called closedness conditions), and they are actually equivalent to each other, as well as quivalent to an extended version of the so-called Hahn-Banach-Lagrange theorem, and its stable version, correspondingly. It is shown that any of them implies analytic and algebraic versions of the Hahn-Banach theorem and the Mazur-Orlicz theorem for extended sublinear functions.
Full stability of locally optimal solutions in second-order cone programs
- Authors: Mordukhovich, Boris , Outrata, Jiri , Sarabi, Ebrahim
- Date: 2014
- Type: Text , Journal article
- Relation: SIAM Journal on Optimization Vol. 24, no. 4 (2014), p. 1581-1613
- Full Text:
- Reviewed:
- Description: The paper presents complete characterizations of Lipschitzian full stability of locally optimal solutions to second-order cone programs (SOCPs) expressed entirely in terms of their initial data. These characterizations are obtained via appropriate versions of the quadratic growth and strong second-order sufficient conditions under the corresponding constraint qualifications. We also establish close relationships between full stability of local minimizers for SOCPs and strong regularity of the associated generalized equations at nondegenerate points. Our approach is mainly based on advanced tools of second-order variational analysis and generalized differentiation.
Gradient-free method for nonsmooth distributed optimization
- Authors: Li, Jueyou , Wu, Changzhi , Wu, Zhiyou , Long, Qiang
- Date: 2014
- Type: Text , Journal article
- Relation: Journal of Global Optimization Vol.61, no.2 (March 2014), p.325-340
- Full Text:
- Reviewed:
- Description: In this paper, we consider a distributed nonsmooth optimization problem over a computational multi-agent network. We first extend the (centralized) Nesterov’s random gradient-free algorithm and Gaussian smoothing technique to the distributed case. Then, the convergence of the algorithm is proved. Furthermore, an explicit convergence rate is given in terms of the network size and topology. Our proposed method is free of gradient, which may be preferred by practical engineers. Since only the cost function value is required, our method may suffer a factor up to d (the dimension of the agent) in convergence rate over that of the distributed subgradient-based methods in theory. However, our numerical simulations show that for some nonsmooth problems, our method can even achieve better performance than that of subgradient-based methods, which may be caused by the slow convergence in the presence of subgradient.
Metric Regularity of the Sum of Multifunctions and Applications
- Authors: Van Ngai, Huynh , Tron, Nguyen Tron , Thera, Michel
- Date: 2014
- Type: Text , Journal article
- Relation: Journal of Optimization Theory and Applications Vol. 160, no. 2 (2014), p. 355-390
- Relation: http://purl.org/au-research/grants/arc/DP110102011
- Full Text: false
- Reviewed:
- Description: The metric regularity of multifunctions plays a crucial role in modern variational analysis and optimization. This property is a key to study the stability of solutions of generalized equations. Many practical problems lead to generalized equations associated to the sum of multifunctions. This paper is devoted to study the metric regularity of the sum of multifunctions. As the sum of closed multifunctions is not necessarily closed, almost all known results in the literature on the metric regularity for one multifunction (which is assumed usually to be closed) fail to imply regularity properties of the sum of multifunctions. To avoid this difficulty, we use an approach based on the metric regularity of so-called epigraphical multifunctions and the theory of error bounds to study the metric regularity of the sum of two multifunctions, as well as some related important properties of variational systems. Firstly, we establish the metric regularity of the sum of a regular multifunction and a pseudo-Lipschitz multifunction with a suitable Lipschitz modulus. These results subsume some recent results by Durea and Strugariu. Secondly, we derive coderivative characterizations of the metric regularity of epigraphical multifunctions associated with the sum of multifunctions. Applications to the study of the behavior of solutions of variational systems are reported. © 2013 Springer Science+Business Media New York.
On optimal control of a sweeping process coupled with an ordinary differential equation
- Authors: Adam, Lukas , Outrata, Jiri
- Date: 2014
- Type: Text , Journal article
- Relation: Discrete and Continuous Dynamical Systems - Series B Vol. 19, no. 9 (November 2014 2014), p. 2709-2738
- Full Text: false
- Reviewed:
- Description: We study a special case of an optimal control problem governed by a differential equation and a differential rate{independent variational inequality, both with given initial conditions. Under certain conditions, the variational inequality can be reformulated as a differential inclusion with discontinuous right-hand side. This inclusion is known as sweeping process. We perform a discretization scheme and prove the convergence of optimal solutions of the discretized problems to the optimal solution of the original problem. For the discretized problems we study the properties of the solution map and compute its coderivative. Employing an appropriate chain rule, this enables us to compute the subdifferential of the objective function and to apply a suitable optimization technique to solve the discretized problems. The investigated problem is used to model a situation arising in the area of queuing theory.
On relaxing the Mangasarian-Fromovitz constraint qualification
- Authors: Kruger, Alexander , Minchenko, Leonld , Outrata, Jiri
- Date: 2014
- Type: Text , Journal article
- Relation: Positivity Vol. 18, no. 1 (2014), p. 171-189
- Relation: http://purl.org/au-research/grants/arc/DP110102011
- Full Text:
- Reviewed:
- Description: For the classical nonlinear program, two new relaxations of the Mangasarian– Fromovitz constraint qualification are discussed and their relationship with some standard constraint qualifications is examined. In particular, we establish the equivalence of one of these constraint qualifications with the recently suggested by Andreani et al. Constant rank of the subspace component constraint qualification. As an application, we make use of this new constraint qualification in the local analysis of the solution map to a parameterized equilibrium problem, modeled by a generalized equation.
On topological existence theorems and applications to optimization-related problems
- Authors: Khanh, Phan Quoc , Lin, Lai Jiu , Long, Vo Si Trong
- Date: 2014
- Type: Text , Journal article
- Relation: Mathematical Methods of Operations Research Vol. 79, no. 3 (June 2014 2014), p. 253-272
- Full Text: false
- Reviewed:
- Description: In this paper, we establish a continuous selection theorem and use it to derive five equivalent results on the existence of fixed points, sectional points, maximal elements, intersection points and solutions of variational relations, all in topological settings without linear structures. Then, we study the solution existence of a number of optimization-related problems as examples of applications of these results: quasivariational inclusions, Stampacchia-type vector equilibrium problems, Nash equilibria, traffic networks, saddle points, constrained minimization, and abstract economies.
- Description: C1
Optimality conditions and optimization methods for quartic polynomial optimization
- Authors: Wu, Zhiyou , Tian, Jing , Quan, Jing , Ugon, Julien
- Date: 2014
- Type: Text , Journal article
- Relation: Applied Mathematics and Computation Vol. 232, no. (2014), p. 968-982
- Full Text: false
- Reviewed:
- Description: In this paper multivariate quartic polynomial optimization program (QPOP) is considered. Quartic optimization problems arise in various practical applications and are proved to be NP hard. We discuss necessary global optimality conditions for quartic problem (QPOP). And then we present a new (strongly or ε-strongly) local optimization method according to necessary global optimality conditions, which may escape and improve some KKT points. Finally we design a global optimization method for problem (QPOP) by combining the new (strongly or ε-strongly) local optimization method and an auxiliary function. Numerical examples show that our algorithms are efficient and stable.
Post-buckling solutions of hyper-elastic beam by canonical dual finite element method
- Authors: Cai, Kun , Gao, David , Qin, Qing
- Date: 2014
- Type: Text , Journal article
- Relation: Mathematics and Mechanics of Solids Vol. 19, no. 6 (2014), p. 659-671
- Full Text: false
- Reviewed:
- Description: The post-buckling problem of a large deformed beam is analyzed using the canonical dual finite element method (CD-FEM). The feature of this method is to choose correctly the canonical dual stress so that the original non-convex potential energy functional is reformulated in a mixed complementary energy form with both displacement and stress fields, and a pure complementary energy is explicitly formulated in finite dimensional space. Based on the canonical duality theory and the associated triality theorem, a primal–dual algorithm is proposed, which can be used to find all possible solutions of this non-convex post-buckling problem. Numerical results show that the global maximum of the pure-complementary energy leads to a stable buckled configuration of the beam, while the local extrema of the pure-complementary energy present unstable deformation states. We discovered that the unstable buckled state is very sensitive to the number of total elements and the external loads. Theoretical results are verified through numerical examples and some interesting phenomena in post-bifurcation of this large deformed beam are observed.
Preface of the special issue OR: Connecting sciences supported by global optimization related to the 25th European conference on operational research (EURO XXV 2012)
- Authors: Bagirov, Adil , Miettinen, Kaisa , Weber, Gerhard-Wilhelm
- Date: 2014
- Type: Text , Journal article
- Relation: Journal of Global Optimization Vol. 60, no. 1 (June 2014), p. 1-3
- Full Text: false
- Reviewed:
- Description: C1
Shape optimization in contact problems with Coulomb friction and a solution-dependent friction coefficient
- Authors: Beremlijski, Petr , Haslinger, Jaroslav , Outrata, Jiri , Pathó, Róbert
- Date: 2014
- Type: Text , Journal article
- Relation: SIAM Journal on Control and Optimization Vol. 52, no. 5 (2014), p. 3371-3400
- Full Text: false
- Reviewed:
- Description: The present paper deals with shape optimization in discretized two-dimensional (2D) contact problems with Coulomb friction, where the coefficient of friction is assumed to depend on the unknown solution. Discretization of the continuous state problem leads to a system of finite-dimensional implicit variational inequalities, parametrized by the so-called design variable, that determines the shape of the underlying domain. It is shown that if the coefficient of friction is Lipschitz and sufficiently small in the C0,1 -norm, then the discrete state problems are uniquely solvable for all admissible values of the design variable (the admissible set is assumed to be compact), and the state variables are Lipschitzian functions of the design variable. This facilitates the numerical solution of the discretized shape optimization problem by the so-called implicit programming approach. Our main results concern sensitivity analysis, which is based on the well-developed generalized differential calculus of B. Mordukhovich and generalizes some of the results obtained in this context so far. The derived subgradient information is then combined with the bundle trust method to compute several model examples, demonstrating the applicability and efficiency of the presented approach. © 2014 Society for Industrial and Applied Mathematics
Sigma supporting cone and optimality conditions in non-convex problems
- Authors: Hassani, Sara , Mammadov, Musa
- Date: 2014
- Type: Text , Journal article
- Relation: Far East Journal of Mathematical Sciences Vol. 91, no. 2 (2014), p. 169-190
- Full Text: false
- Reviewed:
- Description: In this paper, a new supporting function for characterizing non-convex sets is introduced. The notions of σ-supporting cone and maximal conic gap are proposed and some properties are investigated. By applying these new notions, we establish the optimality conditions considered in [7] for a broader class of finite dimensional normed spaces in terms of weak subdifferentials.
Solving second-order conic systems with variable precision
- Authors: Cucker, Felipe , Peña, Javier , Roshchina, Vera
- Date: 2014
- Type: Text , Journal article
- Relation: Mathematical Programming Vol. 150, no. 2 (2014), p. 217-250
- Full Text: false
- Reviewed:
- Description: We describe and analyze an interior-point method to decide feasibility problems of second-order conic systems. A main feature of our algorithm is that arithmetic operations are performed with finite precision. Bounds for both the number of arithmetic operations and the finest precision required are exhibited. © 2014, Springer-Verlag Berlin Heidelberg and Mathematical Optimization Society.
Some preconditioners for systems of linear inequalities
- Authors: Peña, Javier , Roshchina, Vera , Soheili, Negar
- Date: 2014
- Type: Text , Journal article
- Relation: Optimization Letters Vol. 8, no. 7 (2014), p. 2145-2152
- Full Text: false
- Reviewed:
- Description: We show that a combination of two simple preprocessing steps would generally improve the conditioning of a homogeneous system of linear inequalities. Our approach is based on a comparison among three different but related notions of conditioning for linear inequalities. © 2014, Springer-Verlag Berlin Heidelberg.
Special Issue on recent advances in continuous optimization on the occasion of the 25th European conference on Operational Research (EURO XXV 2012)
- Authors: Weber, Gerhard-Wilhelm , Kruger, Alexander , Martinez-Legaz, Juan , Mordukhovich, Boris , Sakalauskas, Leonidas
- Date: 2014
- Type: Text , Journal article
- Relation: Optimization Vol. 63, no. 1 (2014), p. 1-5
- Full Text:
- Reviewed:
Spline regression models for complex multi-modal regulatory networks
- Authors: Ozmen, Ayse , Kropat, Erik , Weber, Gerhard-Wilhelm
- Date: 2014
- Type: Text , Journal article
- Relation: Optimization Methods and Software Vol. 29, no. 3 (2014), p. 515-534
- Full Text: false
- Reviewed:
- Description: Complex regulatory networks often have to be further expanded and improved with regard to the unknown effects of additional parameters and factors that can emit a disturbing influence on the key variables under consideration. The concept of target-environment (TE) networks provides a holistic framework for the analysis of such parameter-dependent multi-modal systems. In this study, we consider time-discrete TE regulatory systems with spline entries. We introduce a new regression model for these particular two-modal systems that allows us to determine the unknown system parameters by applying the multivariate adaptive regression spline (MARS) technique and the newly developed conic multivariate adaptive regression spline (CMARS) method. We obtain a relaxation by means of continuous optimization, especially, conic quadratic programming (CQP) that could be conducted by interior point methods. Finally, a numerical example demonstrates the efficiency of the spline-based approach.
Structure learning of Bayesian Networks using global optimization with applications in data classification
- Authors: Taheri, Sona , Mammadov, Musa
- Date: 2014
- Type: Text , Journal article
- Relation: Optimization Letters Vol. 9, no. 5 (2014), p. 931-948
- Full Text:
- Reviewed:
- Description: Bayesian Networks are increasingly popular methods of modeling uncertainty in artificial intelligence and machine learning. A Bayesian Network consists of a directed acyclic graph in which each node represents a variable and each arc represents probabilistic dependency between two variables. Constructing a Bayesian Network from data is a learning process that consists of two steps: learning structure and learning parameter. Learning a network structure from data is the most difficult task in this process. This paper presents a new algorithm for constructing an optimal structure for Bayesian Networks based on optimization. The algorithm has two major parts. First, we define an optimization model to find the better network graphs. Then, we apply an optimization approach for removing possible cycles from the directed graphs obtained in the first part which is the first of its kind in the literature. The main advantage of the proposed method is that the maximal number of parents for variables is not fixed a priory and it is defined during the optimization procedure. It also considers all networks including cyclic ones and then choose a best structure by applying a global optimization method. To show the efficiency of the algorithm, several closely related algorithms including unrestricted dependency Bayesian Network algorithm, as well as, benchmarks algorithms SVM and C4.5 are employed for comparison. We apply these algorithms on data classification; data sets are taken from the UCI machine learning repository and the LIBSVM. © 2014, Springer-Verlag Berlin Heidelberg.