Unconstrained optimization based on homogeneous models

In this paper a generalized algorithm for unconstrained optimization based on homogeneous functions is presented. It is shown that this generalized algorithm reaches the minimum of a homogeneous function inn + 2 iterations and does not require finding the minimum on the one-dimensional search. The Jacobson and Oksman algorithm can be derived from this general algorithm.This work was supported by the National Research Council of Canada under grant A 7239 and by a scholarship to the author.     

Parallel arc-allocation algorithms for optimizing generalized networks

Two parallel shared-memory algorithms are presented for the optimization of generalized networks. These algorithms are based on the allocation of arc-related operations in the (generalized) network simplex method. One method takes advantage of the multi-tree structure of basic solutions and performs pivot operations in parallel, utilizing locking to ensure correctness. The other algorithm utilizes only one processor for sequential pivoting, but parallelizes the pricing operation and overlaps this task with pivoting in a speculative manner (i.e. since pivoting and pricing involve data dependencies, a candidate for flow change generated by the pricing processors is not guaranteed to be acceptable, but in practice generally has this property). The relative performance of these two methods (on the Sequent Symmetry S81 multiprocessor) is compared and contrasted with that of a fast sequential algorithm on a set of large-scale test problems of up to 1,000,000 arcs.This research was supported in part by NSF grant CCR-8709952 and AFOSR grant AFOSR-86-0194.     

Viscosity approximations by generalized contractions for resolvents of accretive operators in Banach spaces

In this paper, we prove a strong convergence theorem for resolvents of accretive operators in a Banach space by the viscosity approximation method with a generalized contraction mapping. The proximal point algorithm in a Banach space is also considered. The results extend some very recent theorems of W. Takahashi.     

Multidimensional greatest common divisor and Lehmer algorithms

A class of multidimensional greatest common divisor algorithms is studied. Their connection with the Jacobi algorithm is established and used to obtain theoretical properties such as the existence of digit frequencies. A technique of D. H. Lehmer's for Euclid's algorithm is generalized for efficient computation of the multidimensional algorithms. For triples of integers, two algorithms of interest are studied empirically.This work was partially supported by NSF grant #DCR75-07070.     

Optimum departure times for commuters in congested networks

We propose an algorithm to compute the optimum departure time and path for a commuter in a congested network. Constant costs for use of arcs, cost functions of travel time depending on exogenous congestion and schedule delay are taken into account. A best path for a given departure time is computed with a previous algorithm for the generalized shortest path problem. The globally optimal departure time and an optimal path are determined by adapting Piyavskii's algorithm to the case of one-sided Lipschitz functions.This research has benefited from a grant of the Transportation Center of Northwestern University. The first author's research was partially supported by NSF grant No. SES-8911517 to Northwestern University. The second author's research was partially supported by AFOSR grants No. 89-0512 and 90-0008 to Rutgers University.     

HGLM modelling of dropout process using a frailty model

Summary  We introduce a shared random-effect model, derived from frailty models to account for informative dropout. We extend the iterative weighted least squares algorithm for hierarchical generalized linear models to shared random-effect models. Monte-Carlo simulations are carried out to illustrate that the proposed method works well whether the random-effect distribution is correctly specified or not. This study was supported by a grant of the Korea Health 21 R & D Project, Ministry of Health & Welfare, Republic of Korea. (01-PJ1-PG3-51200-0002).     

Further applications of a splitting algorithm to decomposition in variational inequalities and convex programming

A classical method for solving the variational inequality problem is the projection algorithm. We show that existing convergence results for this algorithm follow from one given by Gabay for a splitting algorithm for finding a zero of the sum of two maximal monotone operators. Moreover, we extend the projection algorithm to solveany monotone affine variational inequality problem. When applied to linear complementarity problems, we obtain a matrix splitting algorithm that is simple and, for linear/quadratic programs, massively parallelizable. Unlike existing matrix splitting algorithms, this algorithm converges under no additional assumption on the problem. When applied to generalized linear/quadratic programs, we obtain a decomposition method that, unlike existing decomposition methods, can simultaneously dualize the linear constraints and diagonalize the cost function. This method gives rise to highly parallelizable algorithms for solving a problem of deterministic control in discrete time and for computing the orthogonal projection onto the intersection of convex sets.This research is partially supported by the U.S. Army Research Office, contract DAAL03-86-K-0171 (Center for Intelligent Control Systems), and by the National Science Foundation under grant NSF-ECS-8519058.Thanks are due to Professor J.-S. Pang for his helpful comments.     

On the classification of generalized Fischer spaces

In this paper we show that a generalized Fischer space is either a Fischer space or is locally a polar space. As a corollary we obtain the classification of the finite irreducible generalized Fischer spaces.This work was partially supported by a grant from the National Science Foundation, U.S.A.     

Generalized Jacobian for Functions with Infinite Dimensional Range and Domain

In this paper, locally Lipschitz functions acting between infinite dimensional normed spaces are considered. When the range is a dual space and satisfies the Radon–Nikodym property, Clarke’s generalized Jacobian will be extended to this setting. Characterization and fundamental properties of the extended generalized Jacobian are established including the nonemptiness, the β-compactness, the β-upper semicontinuity, and a mean-value theorem. A connection with known notions is provided and chain rules are proved using key results developed. This included the vectorization and restriction theorem, and the extension theorem. Therefore, the generalized Jacobian introduced in this paper is proved to enjoy all the properties required of a derivative like-set. Research of the first author is supported by the Hungarian Scientific Research Fund (OKTA) under grant K62316. Research of the second author is supported by the National Science Foundation under grant DMS-0306260.     

A Feasibility Pump for mixed integer nonlinear programs

We present an algorithm for finding a feasible solution to a convex mixed integer nonlinear program. This algorithm, called Feasibility Pump, alternates between solving nonlinear programs and mixed integer linear programs. We also discuss how the algorithm can be iterated so as to improve the first solution it finds, as well as its integration within an outer approximation scheme. We report computational results. P. Bonami is supported in part by a grant from IBM and by ANR grant BLAN06-1-138894. G. Cornuéjols is supported in part by NSF grant CMMI-0653419, ANR grant BLAN06-1-138894 and ONR grant N00014-03-1-0188. Part of this research was carried out when Andrea Lodi was Herman Goldstine Fellow of the IBM T.J. Watson Research Center whose support is gratefully acknowledged. F. Margot is supported in part by a grant from IBM and by ONR grant N00014-03-1-0188.     

Nonlinear mappings associated with the generalized linear complementarity problem

We show that the Cottle—Dantzig generalized linear complementarity problem (GLCP) is equivalent to a nonlinear complementarity problem (NLCP), a piecewise linear system of equations (PLS), a multiple objective programming problem (MOP), and a variational inequalities problem (VIP). On the basis of these equivalences, we provide an algorithm for solving problem GLCP.Project partially supported by a grant from Oak Ridge Associated Universities, TN, USA.     

A simplex algorithm for piecewise-linear programming II: Finiteness,feasibility and degeneracy

The simplex method for linear programming can be extended to permit the minimization of any convex separable piecewise-linear objective, subject to linear constraints. Part I of this paper has developed a general and direct simplex algorithm for piecewise-linear programming, under convenient assumptions that guarantee a finite number of basic solutions, existence of basic feasible solutions, and nondegeneracy of all such solutions. Part II now shows how these assumptions can be weakened so that they pose no obstacle to effective use of the piecewise-linear simplex algorithm. The theory of piecewise-linear programming is thereby extended, and numerous features of linear programming are generalized or are seen in a new light. An analysis of the algorithm's computational requirements and a survey of applications will be presented in Part III.This research has been supported in part by the National Science Foundation under grant DMS-8217261.     

Iterative Algorithm for Generalized Set-Valued Strongly Nonlinear Mixed Variational-Like Inequalities

The auxiliary principle technique is extended to study the generalized strongly nonlinear mixed variational-like inequality problem for set-valued mappings without compact values. We establish first the existence of a solution of the related auxiliary problem. Then, the iterative algorithm for solving that problem is given by using this existence result. Moreover, the existence of a solution of the original problem and the convergence of iterative sequences generated by the algorithm are both derived.Research partially supported by the Teaching and Research Award Fund for Outstanding Young Teachers in Higher Education Institutions of MOE, China and the Dawn Program Foundation in Shanghai, China. Research partially supported by a grant from the National Science Council of Taiwan     

Characterizations of linear suboptimality for mathematical programs with equilibrium constraints

The paper is devoted to the study of a new notion of linear suboptimality in constrained mathematical programming. This concept is different from conventional notions of solutions to optimization-related problems, while seems to be natural and significant from the viewpoint of modern variational analysis and applications. In contrast to standard notions, it admits complete characterizations via appropriate constructions of generalized differentiation in nonconvex settings. In this paper we mainly focus on various classes of mathematical programs with equilibrium constraints (MPECs), whose principal role has been well recognized in optimization theory and its applications. Based on robust generalized differential calculus, we derive new results giving pointwise necessary and sufficient conditions for linear suboptimality in general MPECs and its important specifications involving variational and quasivariational inequalities, implicit complementarity problems, etc. Research was partially supported by the National Science Foundation under grant DMS-0304989 and by the Australian Research Council under grant DP-0451168.     

Connectedness of the Set of Efficient Solutions for Generalized Systems

We introduce the concept of positive proper efficient solutions to the generalized system in this paper. We show that, under some suitable conditions, the set of positive proper efficient solutions is dense in the set of efficient solutions to the generalized system. We discuss also the connectedness of the set of efficient solutions for the generalized system with monotone bifunctions in real locally convex Hausdorff topological vector spaces. This research was partially supported by the National Natural Science Foundation of China (10561007), the Natural Science Foundation of Jiangxi Province, China, and a grant from the National Science Council of ROC.     

A Newton method for a class of quasi-variational inequalities

A variant of the Newton method for nonsmooth equations is applied to solve numerically quasivariational inequalities with monotone operators. For this purpose, we investigate the semismoothness of a certain locally Lipschitz operator coming from the quasi-variational inequality, and analyse the generalized Jacobian of this operator to ensure local convergence of the method. A simplified variant of this approach, applicable to implicit complementarity problems, is also studied. Small test examples have been computed.This work has been supported in parts by a grant from the German Scientific Foundation and by a grant from the Czech Academy of Sciences.     

Superlinear primal-dual affine scaling algorithms for LCP

We describe an interior-point algorithm for monotone linear complementarity problems in which primal-dual affine scaling is used to generate the search directions. The algorithm is shown to have global and superlinear convergence with Q-order up to (but not including) two. The technique is shown to be consistent with a potential-reduction algorithm, yielding the first potential-reduction algorithm that is both globally and superlinearly convergent.Corresponding author. The work of this author was based on research supported by the Office of Scientific Computing, U.S. Department of Energy, under Contract W-31-109-Eng-38.The work of this author was based on research supported by the National Science Foundation under grant DDM-9109404 and the Office of Naval Research under grant N00014-93-1-0234. This work was done while the author was a faculty member of the Systems and Industrial Engineering Department at the University of Arizona.     

An implicit shift bidiagonalization algorithm for ill-posed systems

Iterative methods based on Lanczos bidiagonalization with full reorthogonalization (LBDR) are considered for solving large-scale discrete ill-posed linear least-squares problems of the form min _x Ax–b₂. Methods for regularization in the Krylov subspaces are discussed which use generalized cross validation (GCV) for determining the regularization parameter. These methods have the advantage that no a priori information about the noise level is required. To improve convergence of the Lanczos process we apply a variant of the implicitly restarted Lanczos algorithm by Sorensen using zero shifts. Although this restarted method simply corresponds to using LBDR with a starting vector (AA ^T) ^p b, it is shown that carrying out the process implicitly is essential for numerical stability. An LBDR algorithm is presented which incorporates implicit restarts to ensure that the global minimum of the CGV curve corresponds to a minimum on the curve for the truncated SVD solution. Numerical results are given comparing the performance of this algorithm with non-restarted LBDR.This work was partially supported by DARPA under grant 60NANB2D1272 and by the National Science Foundation under grant CCR-9209349.     

Asymptotic theory in generalized ilinear models with nuisance scale parameters

Summary This paper establishes the asymptotic normality and the consistencyrobustness of the weighted least squares estimator (WLSE) in the generalized linear models with multiple nuisance scale parameters. In addition, noting that the asymptotic robust statistical inference in presence of nuisance scale parameters requires a consistency-robust estimator of the asymptotic covariance matrix of the WLSE, this paper derives a class of covariance estimators and proves their consistency-robustness.This research was supported by an operating grant from the Natural Science and Engineering Research Council of Canada and the United States NSF-AFOSR grant ISSA-860068     

Approximation numbers and Yamamoto's theorem in Banach algebras

Yamamoto's Theorem is an asymptotic relation between the singular values and eigenvalues of a matrix. There are formulations of this result involving generalized singular values (approximation numbers) of matrices and, more generally, bounded linear operators on Banach spaces. In this paper, we prove Yamamoto type theorems for Banach algebras.This work partially supported by NSF grant DMS 88-02836