ALBERT

All Library Books, journals and Electronic Records Telegrafenberg

Your email was sent successfully. Check your inbox.

An error occurred while sending the email. Please try again.

Proceed reservation?

Export
Filter
  • Articles  (53)
  • convergence  (53)
  • Springer  (53)
  • American Geophysical Union (AGU)
  • Annual Reviews
  • 2015-2019
  • 1995-1999  (41)
  • 1980-1984  (12)
  • 1935-1939
  • Mathematics  (53)
  • 1
    Electronic Resource
    Electronic Resource
    Springer
    Monatshefte für Mathematik 125 (1998), S. 1-14 
    ISSN: 1436-5081
    Keywords: 11J70 ; 11K50 ; 28D05 ; Continued fractions ; expansion ; generators ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The concept of convergence of continued fraction type algorithms has been defined a number of times in the literature. We investigate the relation between these definitions, and show that they do not always coincide. We relate the definitions to the question whether or not the natural partition of the underlying dynamical system is a generator. It turns out that the ‘right’ definition of convergence is equivalent to this partition being a generator. The second definition of convergence is shown to be equivalent only under extra conditions on the transformation. These extra conditions are typically found to be satisfied when the second definition is used in the literature.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 2
    Electronic Resource
    Electronic Resource
    Springer
    Journal of dynamics and differential equations 7 (1995), S. 549-566 
    ISSN: 1572-9222
    Keywords: Stochastic equations ; approximate inertial manifolds ; exponential order ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract A class of semilinear nonautonomous parabolic equations subjected to additive white noise is considered. The existence of a family of randomN-dimensional approximate inertial manifolds (AIMs) whose neighborhoods of thickness of order exp(-δλ N+1 β ) attract exponentially in the mean all the trajectories is proved forN large enough. Hereλ N+1 is the (N+1)th eigenvalue of the corresponding linear problem, andδ andΒ are positive constants. We also construct a sequence of AIMs which converges to the exact inertial manifold, when a spectral gap condition is satisfied. These results remain true for deterministic autonomous and nonautonomous cases.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 3
    Electronic Resource
    Electronic Resource
    Springer
    Journal of dynamics and differential equations 9 (1997), S. 133-155 
    ISSN: 1572-9222
    Keywords: Wave equation ; dissipation ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract We prove that any bounded solution (u, u 1) ofu u +du t −Δu+f(u)=0,u=u(x, t), x∈ℝN,N⩾3, converges to a fixed stationary state provided its initial energy is appropriately small. The theory of concentrated compactness is used in combination with some recent results concerning the uniqueness of the so-called ground-state solution of the corresponding stationary problem.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 4
    Electronic Resource
    Electronic Resource
    Springer
    Annals of operations research 56 (1995), S. 189-208 
    ISSN: 1572-9338
    Keywords: Epi-distance ; large deviations ; stochastic programs ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics , Economics
    Notes: Abstract Several exponential bounds are derived by means of the theory of large deviations for the convergence of approximate solutions of stochastic optimization problems. The basic results show that the solutions obtained by replacing the original distribution by an empirical distribution provides an effective tool for solving stochastic programming problems.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 5
    Electronic Resource
    Electronic Resource
    Springer
    Annals of operations research 56 (1995), S. 251-285 
    ISSN: 1572-9338
    Keywords: Stochastic programming ; quadratic programming ; convergence ; B-regularity
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics , Economics
    Notes: Abstract Extended Linear-Quadratic Programming (ELQP) problems were introduced by Rockafellar and Wets for various models in stochastic programming and multistage optimization. Several numerical methods with linear convergence rates have been developed for solving fully quadratic ELQP problems, where the primal and dual coefficient matrices are positive definite. We present a two-stage sequential quadratic programming (SQP) method for solving ELQP problems arising in stochastic programming. The first stage algorithm realizes global convergence and the second stage algorithm realizes superlinear local convergence under a condition calledB-regularity.B-regularity is milder than the fully quadratic condition; the primal coefficient matrix need not be positive definite. Numerical tests are given to demonstrate the efficiency of the algorithm. Solution properties of the ELQP problem underB-regularity are also discussed.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 6
    Electronic Resource
    Electronic Resource
    Springer
    BIT 35 (1995), S. 516-539 
    ISSN: 1572-9125
    Keywords: Nonsymmetric eigenvalue problem ; generalized block Lanczos ; block Arnoldi ; convergence ; banded block Hessenberg matrix
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The incomplete orthogonalization method (IOM) proposed by Saad for computing a few eigenpairs of large nonsymmetric matrices is generalized into a block incomplete orthogonalization method (BIOM). It is studied how the departure from symmetry ‖A − A H ‖ affects the conditioning of the block basis vectors generated by BIOM, and some relationships are established between the approximate eigenpairs obtained by BIOM and Ritz pairs. It is proved that BIOM behaves much like generalized block Lanczos methods if the basis vectors of the block Krylov subspace generated by it are strongly linearly independent. However, it is shown that BIOM may generate a nearly linearly dependent basis for a general nonsymmetric matrix. Numerical experiments illustrate the convergence behavior of BIOM.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 7
    Electronic Resource
    Electronic Resource
    Springer
    BIT 36 (1996), S. 635-652 
    ISSN: 1572-9125
    Keywords: Stiff differential equations ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Existing convergence concepts for the analysis of discretizations of nonlinear stiff problems suffer from considerable drawbacks. Our intention is to extend the convergence theory to a relevant class of nonlinear problems, where stiffness is axiomatically characterized in natural geometric terms. Our results will be presented in a series of papers. In the present paper (Part I) we motivate the need for such an extension of the existing theory, and our approach is illustrated by means of a convergence argument for the Implicit Euler scheme.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 8
    Electronic Resource
    Electronic Resource
    Springer
    BIT 21 (1981), S. 232-241 
    ISSN: 1572-9125
    Keywords: Ams (Mos): 65R20 ; CR: 5.18 ; Numerical analysis ; Volterra integral equations of the first kind ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Quadrature rules, generated by linear multistep methods for ordinary differential equations, are employed to construct a wide class of direct quadrature methods for the numerical solution of first kind Volterra integral equations. Our class covers several methods previously considered in the literature. The methods are convergent provided that both the first and second characteristic polynomial of the linear multistep method satisfy the root condition. Furthermore, the stability behaviour for fixed positive values of the stepsizeh is analyzed, and it turns out that convergence implies (fixedh) stability. The subclass formed by the backward differentiation methods up to order six is discussed and illustrated with numerical examples.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 9
    Electronic Resource
    Electronic Resource
    Springer
    Acta applicandae mathematicae 47 (1997), S. 213-220 
    ISSN: 1572-9036
    Keywords: convergence ; convergence rate ; convergence order ; sequences
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Eight measures of rate of convergence of monotone sequences of real numbers have been proposed and discussed in a paper by Beyer, Ebanks, and Qualls (Acta Appl. Math. 20 (1990), 267–284). These rates arose in discussions of dynamical systems. In the comparisons of these rates, two problems remained. A complete answer is given to one of the problems and a partial answer is given to the other problem.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 10
    Electronic Resource
    Electronic Resource
    Springer
    Advances in computational mathematics 3 (1995), S. 197-218 
    ISSN: 1572-9044
    Keywords: Conservation laws ; higher order accurate finite volume methods ; irregular grids ; convergence ; 35L65 ; 65M12 ; 65M50
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract We prove convergence to the entropy solution of a general class of higher order finite volume schemes on unstructured, irregular grids for multidimensional scalar conservation laws. Such grids allow for cells to become flat in the limit. We derive a new entropy inequality for higher order schemes built on Godunov's numerical flux. Our result implies convergence of suitably modified versions of MUSCL-type finite volume schemes, ENO schemes and the discontinuous Galerkin finite element method.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 11
    Electronic Resource
    Electronic Resource
    Springer
    Czechoslovak mathematical journal 49 (1999), S. 767-777 
    ISSN: 1572-9141
    Keywords: variational inclusion ; fuzzy mapping ; algorithm ; existence ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In this paper, we introduce and study a new class of completely generalized nonlinear variational inclusions for fuzzy mappings and construct some new iterative algorithms. We prove the existence of solutions for this kind of completely generalized nonlinear variational inclusions and the convergence of iterative sequences generated by the algorithms.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 12
    Electronic Resource
    Electronic Resource
    Springer
    Extremes 2 (1999), S. 295-307 
    ISSN: 1572-915X
    Keywords: Shock ; intershock time ; stopped random walk ; renewal theory ; first passage times ; moments ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The standard assumptions in shock models are that the failure (of the system) is related either to the cumulative effect of a (large) number of shocks or that failure is caused by a shock which is larger than a certain critical level. The present paper is devoted to the second kind. Here the standard setting is that the shocks Xk, k ≥ 1, and the times between the shocks Yk, k ≥ 1, are independent, identically distributed random vectors (Xk, Yk), k ≥ 1. In particular, Xk and Yk may well be dependent (the typical case). The main object of interest is the time to failure, Tτ(t), where Tn = ∑k≤n Yk and τ(t) is the first exceedance time, viz. the first time that Xk 〉 t. We derive moment relations and asymptotic distributions of Tτ(t) as t increases in such a way that P{X1} 〉 t} tends to 0. A final section discusses some extensions; more general events of failure, the non-i.i.d. case, and point process convergence for a particular case.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 13
    Electronic Resource
    Electronic Resource
    Springer
    Applied mathematics & optimization 7 (1981), S. 1-9 
    ISSN: 1432-0606
    Keywords: Nonlinear programming ; nonconvex programming ; concave minimization ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract A modification of Tuy's cone splitting algorithm for minimizing a concave function subject to linear inequality constraints is shown to be convergent by demonstrating that the limit of a sequence of constructed convex polytopes contains the feasible region. No geometric tolerance parameters are required.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 14
    Electronic Resource
    Electronic Resource
    Springer
    Journal of mathematical biology 20 (1984), S. 159-169 
    ISSN: 1432-1416
    Keywords: Selection ; mating system ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Biology , Mathematics
    Notes: Abstract For a biallelic model of differential self-fertilization and differential positive assortative mating based on genotype, it is shown that the genotypic frequencies converge for all sets of mating system parameters. Overdominance and underdominance with respect to the parameters are necessary but not sufficient conditions for global convergence to a polymorphic equilibrium and local attractiveness of both the fixation states, respectively. There are cases of overdominance and underdominance for which one fixation state is globally attractive. The relationship of the result to those known from the classical viability selection model are briefly discussed. For the multiallelic version, it is shown that after the first generation all of the homozygote frequencies are always in excess of the corresponding Hardy-Weinberg proportions if at least one homozygote rate of self-fertilization or assortment probability is positive.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 15
    Electronic Resource
    Electronic Resource
    Springer
    Acta mathematicae applicatae sinica 13 (1997), S. 176-187 
    ISSN: 1618-3932
    Keywords: Spherical surface ; pseudospectral method ; vorticity equations ; stability ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The pseudospectral method for solving vorticity equations on spherical surface is discussed. An interpolation procedure, which is different from the usual ones, is proposed. Based on such an interpolation, the pseudospectral scheme is constructed. Its generalized stability and convergence are analyzed rigorously. The theoretical analysis and computational skills can also be applied to other nonlinear partial differential equations defined on spherical surface.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 16
    Electronic Resource
    Electronic Resource
    Springer
    Acta mathematicae applicatae sinica 11 (1995), S. 17-29 
    ISSN: 1618-3932
    Keywords: Domain Decomposition ; trace average operator ; preconditioner ; convergence ; condition number
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Recently, Bourgat et al.[3] gave a domain decomposition algorithm which can be implemented in parallel, and many numerical experiments have illustrated its efficiency. In this paper, we make a detailed theoretical analysis about this algorithm.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 17
    Electronic Resource
    Electronic Resource
    Springer
    Numerical algorithms 13 (1996), S. 365-398 
    ISSN: 1572-9265
    Keywords: convergence ; multilevel additive methods ; unstructured meshes ; 65N30 ; 65F10
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract We develop a convergence theory for two level and multilevel additive Schwarz domain decomposition methods for elliptic and parabolic problems on general unstructured meshes in two and three dimensions. The coarse and fine grids are assumed only to be shape regular, and the domains formed by the coarse and fine grids need not be identical. In this general setting, our convergence theory leads to completely local bounds for the condition numbers of two level additive Schwarz methods, which imply that these condition numbers are optimal, or independent of fine and coarse mesh sizes and subdomain sizes if the overlap amount of a subdomain with its neighbors varies proportionally to the subdomain size. In particular, we will show that additive Schwarz algorithms are still very efficient for nonselfadjoint parabolic problems with only symmetric, positive definite solvers both for local subproblems and for the global coarse problem. These conclusions for elliptic and parabolic problems improve our earlier results in [12, 15, 16]. Finally, the convergence theory is applied to multilevel additive Schwarz algorithms. Under some very weak assumptions on the fine mesh and coarser meshes, e.g., no requirements on the relation between neighboring coarse level meshes, we are able to derive a condition number bound of the orderO(ρ2 L 2), whereρ = max1≤l≤L(h l +l− 1)/δ l,h l is the element size of thelth level mesh,δ l the overlap of subdomains on thelth level mesh, andL the number of mesh levels.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 18
    Electronic Resource
    Electronic Resource
    Springer
    Numerical algorithms 12 (1996), S. 347-368 
    ISSN: 1572-9265
    Keywords: Continuous two-step Runge-Kutta method ; convergence ; order and stage order ; 65L05 ; 65L06
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract A general class of variable stepsize continuous two-step Runge-Kutta methods is investigated. These methods depend on stage values at two consecutive steps. The general convergence and order criteria are derived and examples of methods of orderp and stage orderq=p orq=p−1 are given forp≤5. Numerical examples are presented which demonstrate that high order and high stage order are preserved on nonuniform meshes with large variations in ratios between consecutive stepsizes.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 19
    Electronic Resource
    Electronic Resource
    Springer
    Numerical algorithms 13 (1996), S. 265-306 
    ISSN: 1572-9265
    Keywords: polar form ; blossom ; Bézier-De Casteljau curve ; patch ; control polygons ; “degree elevation” ; length elevation ; length and width elevation ; height elevation ; convergence ; bounds ; accuracy control ; 65D17 ; 41A10
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract It is known that the sequence of control polygons of a Bézier-De Casteljau curve or surface obtained by the “degree elevation” process converges towards the underlying curve and surface. The notion of blossoming or polar form associated with a polynomial allows to control the accuracy of this convergence and, as a by-product, to give a new and completer proof of convergence.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 20
    Electronic Resource
    Electronic Resource
    Springer
    Numerical algorithms 14 (1997), S. 227-260 
    ISSN: 1572-9265
    Keywords: error control ; convergence ; 34C35 ; 34D05 ; 65L07 ; 65L20 ; 65L50
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract The numerical solution of initial value problems for ordinary differential equations is frequently performed by means of adaptive algorithms with user-input tolerance τ. The time-step is then chosen according to an estimate, based on small time-step heuristics, designed to try and ensure that an approximation to the local error commited is bounded by τ. A question of natural interest is to determine how the global error behaves with respect to the tolerance τ. This has obvious practical interest and also leads to an interesting problem in mathematical analysis. The primary difficulties arising in the analysis are that: (i) the time-step selection mechanisms used in practice are discontinuous as functions of the specified data; (ii) the small time-step heuristics underlying the control of the local error can break down in some cases. In this paper an analysis is presented which incorporates these two difficulties. For a mathematical model of an error per unit step or error per step adaptive Runge–Kutta algorithm, it may be shown that in a certain probabilistic sense, with respect to a measure on the space of initial data, the small time-step heuristics are valid with probability one, leading to a probabilistic convergence result for the global error as τ→0. The probabilistic approach is only valid in dimension m〉1 this observation is consistent with recent analysis concerning the existence of spurious steady solutions of software codes which highlights the difference between the cases m=1 and m〉1. The breakdown of the small time-step heuristics can be circumvented by making minor modifications to the algorithm, leading to a deterministic convergence proof for the global error of such algorithms as τ→0. An underlying theory is developed and the deterministic and probabilistic convergence results proved as particular applications of this theory.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 21
    Electronic Resource
    Electronic Resource
    Springer
    Numerical algorithms 20 (1999), S. 63-73 
    ISSN: 1572-9265
    Keywords: bracketing methods ; bisection ; convergence ; 65H05
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract The computational cost of a bracketing algorithm in the bit model of computation is analyzed, when working with a finite arithmetic of unbounded accuracy. The complexity measure used here is the number of bit operations, seen as a function of the required absolute error of the result. In this model the convergence of the classical bisection method (as well as that of any bracketing method which requires the function sign) is not ensured when no information on the behaviour of the function is available. A modified bisection algorithm with guaranteed convergence is proposed and an upper bound to its computational cost is given.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 22
    Electronic Resource
    Electronic Resource
    Springer
    Acta mathematicae applicatae sinica 13 (1997), S. 337-341 
    ISSN: 1618-3932
    Keywords: Multiplicative iterative algorithm ; inexact line search ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract This paper studies the convergence properties of multiplicative iterative algorithms with inexact line search. We prove that the convergence can be guaranteed for a general form of line search rule, under the assumption of convexity of objective function or the assumption of convergence of the sequence generated by the algorithm. This answers an open problem put forward by Iusem.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 23
    Electronic Resource
    Electronic Resource
    Springer
    Acta mathematicae applicatae sinica 12 (1996), S. 50-58 
    ISSN: 1618-3932
    Keywords: Finite element analysis ; error bounds ; convergence ; eigenpairs ; free vibration ; axisymmetric ; mode number
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Error bounds are obtained for the finite element analysis of the eigenpairs for the free vibration of liquid-filled axisymmetric tanks, for a circumferential mode numberm=1, antisymmetric. The system of finite element equations concerned is derived by means of Galerkin formulation. Finite element numerical results of the eigenvalue problems for the free vibration of the liquid-filled axisymmetric tanks are compared favorably with the theoretical predictions.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 24
    Electronic Resource
    Electronic Resource
    Springer
    Applied mathematics and mechanics 18 (1997), S. 1211-1218 
    ISSN: 1573-2754
    Keywords: shortage ; oscillation ; shock absorber ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mechanical Engineering, Materials Science, Production Engineering, Mining and Metallurgy, Traffic Engineering, Precision Mechanics , Mathematics , Physics
    Notes: Abstract In this paper, an oscillated model, which results from the shortage action in market economy with elastic replacement of goods, is obtained. And some natural relations between the model and a typically oscillatory model are established. The results can interpret some market phenomena and provide the theoretical tools for the economic program.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 25
    Electronic Resource
    Electronic Resource
    Springer
    Applied mathematics and mechanics 19 (1998), S. 1211-1216 
    ISSN: 1573-2754
    Keywords: finite element ; nonconforming element ; axisymmetric ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mechanical Engineering, Materials Science, Production Engineering, Mining and Metallurgy, Traffic Engineering, Precision Mechanics , Mathematics , Physics
    Notes: Abstract A class of modified Wilson arbitrary quadrilateral nonconforming elements for an axisymmetric problem is proposed. Their convergence is proven by means of the strong patch test. The structure of this finite element class is investigated. Thus a general method of axisymmetric nonconforming elements with convergence properties is presented.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 26
    Electronic Resource
    Electronic Resource
    Springer
    Applied mathematics and mechanics 17 (1996), S. 269-274 
    ISSN: 1573-2754
    Keywords: junction ; scaling ; junction condition ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mechanical Engineering, Materials Science, Production Engineering, Mining and Metallurgy, Traffic Engineering, Precision Mechanics , Mathematics , Physics
    Notes: Abstract In this paper, a kind of junction problem between linear elastomer and thin plate is changed, through a well scaling, into an equivalent problem which is studied easily. Then, using an idea of disturbance, we have studied the convergence of the displacement vector field under the action of body force, which is depending on the thickness of the plate. At last, the variational equations of the limit are obtained.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 27
    Electronic Resource
    Electronic Resource
    Springer
    Lithuanian mathematical journal 39 (1999), S. 20-32 
    ISSN: 1573-8825
    Keywords: Schrödinger equation ; explicit finite-difference schemes ; stability ; energy conservation ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract We consider three-level explicit schemes for solving the nonlinear variable coefficient Schrödinger-type equation. Using spectral and energy methods we establish the stability and convergence of these schemes. The existence of discrete conservation laws is investigated. General results are applied for the DuFort-Frankel and leap-frog diffenrence schemes.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 28
    Electronic Resource
    Electronic Resource
    Springer
    Mathematical notes 61 (1997), S. 100-104 
    ISSN: 1573-8876
    Keywords: series expansions ; spaces of entire functions ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract This paper is devoted to the problem of representing entire functions, in spaces described by the order and the type of these functions, by Lagrange series that converge in the natural topology in these spaces; this topology is stronger than the topology of compact convergence.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 29
    Electronic Resource
    Electronic Resource
    Springer
    Mathematical notes 61 (1997), S. 272-277 
    ISSN: 1573-8876
    Keywords: linear-fractional transformations ; fixed points ; infinite compositions ; convergence ; continued fractions
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract We study the asymptotic behavior of the compositions (Sn o...o S1)(z) and (S1 o...o Sn)(z) of linear-fractional transformations Sn (z) (n=1,2,...) whose fixed points have limits. In particular, if S n (z)=α n (β n +z)-1, then the sequency of compositions (S1o...o Sn)(z) at the point z=0 coincides with the sequence of convergents of the formal continued fraction $$\frac{{\alpha _1 }}{{\beta _1 + \frac{{\alpha _2 }}{{\beta _2 + \cdot \cdot \cdot }}}}.$$ The result obtained can be applied in the study of convergence of formal continued fractions.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 30
    Electronic Resource
    Electronic Resource
    Springer
    Mathematical notes 63 (1998), S. 271-272 
    ISSN: 1573-8876
    Keywords: parabola method ; roots of complex polynomials ; convergence ; iterations
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 31
    Electronic Resource
    Electronic Resource
    Springer
    Mathematical notes 63 (1998), S. 631-637 
    ISSN: 1573-8876
    Keywords: orthonormal system ; Fourier series ; convergence ; divergence ; separation
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract There exist two orthonormal systems such that the Fourier series of each functionƒ ∈ L[0, 1],ƒ ≠ 0, with respect to at least one of these systems diverges on a set of positive measure.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 32
    Electronic Resource
    Electronic Resource
    Springer
    Statistics and computing 6 (1996), S. 337-346 
    ISSN: 1573-1375
    Keywords: Bayesian inference ; convergence ; Gibbs sampling ; longitudinal data ; Markov chain ; Monte Carlo ; parametric transformation ; random effect ; Rubella antibody
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract Bayesian random effects models may be fitted using Gibbs sampling, but the Gibbs sampler can be slow mixing due to what might be regarded as lack of model identifiability. This slow mixing substantially increases the number of iterations required during Gibbs sampling. We present an analysis of data on immunity after Rubella vaccinations which results in a slow-mixing Gibbs sampler. We show that this problem of slow mixing can be resolved by transforming the random effects and then, if desired, expressing their joint prior distribution as a sequence of univariate conditional distributions. The resulting analysis shows that the decline in antibodies after Rubella vaccination is relatively shallow compared to the decline in antibodies which has been shown after Hepatitis B vaccination.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 33
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 30 (1980), S. 211-227 
    ISSN: 1573-2878
    Keywords: Nonlinear programming ; decomposition algorithm ; convexity ; convergence ; SUMT algorithm
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Kronsjö's nonlinear generalization (Ref. 1) of Benders' algorithm (Ref. 2) is reviewed. At each iteration, this algorithm produces upper and lower bounds to the true optimum, and the sequence of lower bounds is increasing. The algorithm is modified, so that the sequence of upper bounds is ε-decreasing as well. The two versions are tested numerically using an ALGOL program originally written by Wong (Ref. 3), incorporating the SUMT method (Fiacco and McCormick, Refs. 4 and 5). The two versions are compared against each other, and the problem of the optimal degree of decomposition is considered. Finally, an attempt is made to express the computer time required to solve the test problems as a function of master problem size, number of subproblems, and average subproblem size.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 34
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 30 (1980), S. 535-547 
    ISSN: 1573-2878
    Keywords: Ill-posed problems ; least-squares estimation ; approximation ; parametric representation ; convergence ; nonconvergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract One standard approach to solvingf(x)=b is the minimization of ∥f(x)−b∥2 overx in $$\mathop \mathfrak{X}\limits^ \sim $$ , where $$\mathop \mathfrak{X}\limits^ \sim $$ corresponds to a parametric representation providing sufficiently good approximation to the true solutionx*. Call the minimizerx=A( $$\mathop \mathfrak{X}\limits^ \sim $$ ). Take $$\mathop \mathfrak{X}\limits^ \sim $$ = $$\mathfrak{X}$$ N for a sequence { $$\mathfrak{X}$$ N } of subspaces becoming dense, and so determine an approximating sequences {x N ≔A ( $$\mathfrak{X}$$ N )}. It is shown, withf linear and one-to-one, that one need not havex N→x* iff −1 is not continuous.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 35
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 31 (1980), S. 303-329 
    ISSN: 1573-2878
    Keywords: Variable metric methods ; quasi-Newton methods ; optimal control problems ; Banach spaces ; convergence ; exact line searches ; control theory
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The purpose of this paper is to extend a family of variable metric methods, of which the BFGS algorithm (Ref. 1) is a member, into function space, in particular, for the solution of unconstrained optimal control problems. An inexact one-dimensional minimization as suggested by Fletcher (ref. 2) is used. It is shown that, with this stepsize rule and under some mild assumptions, the sequence constructed by this family of methods converges superlinearly for a strictly convex functional defined on a suitable Banach space. This result is shown to remain valid on a Hilbert space and on a Euclidean space under more relaxed assumptions. The BFGS method without line searches is used to solve several standard numerical examples, and excellent performance is observed.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 36
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 32 (1980), S. 211-246 
    ISSN: 1573-2878
    Keywords: Optimal control ; control constraints ; terminal equality constraints ; exact penalty function ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract The presence of control constraints, because they are nondifferentiable in the space of control functions, makes it difficult to cope with terminal equality constraints in optimal control problems. Gradient-projection algorithms, for example, cannot be employed easily. These difficulties are overcome in this paper by employing an exact penalty function to handle the cost and terminal equality constraints and using the control constraints to define the space of permissible search directions in the search-direction subalgorithm. The search-direction subalgorithm is, therefore, more complex than the usual linear program employed in feasible-directions algorithms. The subalgorithm approximately solves a convex optimal control problem to determine the search direction; in the implementable version of the algorithm, the accuracy of the approximation is automatically increased to ensure convergence.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 37
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 32 (1980), S. 345-364 
    ISSN: 1573-2878
    Keywords: Optimal control ; control constraints ; terminal equality constraints ; exact penalty functions ; convergence ; relaxed control
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In Part 1 of this paper, implementable and conceptual versions of an algorithm for optimal control problems with control constraints and terminal equality constraints were presented. It was shown that anyL ∞ accumulation points of control sequences generated by the algorithms satisfy necessary conditions of optimality. Since such accumulation points need not exist, it is shown in this paper that control sequences generated by the algorithms always have accumulation points in the sense of control measure, and these accumulation points satisfy optimality conditions for the corresponding relaxed control problem.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 38
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 42 (1984), S. 467-485 
    ISSN: 1573-2878
    Keywords: Optimal control ; hyperbolic partial differential equations ; Darboux boundary conditions ; strong variational algorithm ; convergence ; relaxed control
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In this paper, we consider a class of optimal control problems involving linear hyperbolic partial differential equations with Darboux boundary conditions. A strong variational algorithm has been obtained for solving this class of optimal control problems in a previous paper by the third and the first authors. It was also shown that anyL ∞ accumulation points of control sequences generated by the algorithm satisfy a necessary condition for optimality. Since such accumulation points need not exist, it is shown in this paper that the control sequences generated by the algorithm always have accumulation points in the sense of control measure, and these accumulation points satisfy a necessary condition for optimality for the corresponding relaxed control problems.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 39
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 92 (1997), S. 581-604 
    ISSN: 1573-2878
    Keywords: Algorithms ; unconstrained optimization ; Lyapunov functions ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract Existing algorithms for solving unconstrained optimization problems are generally only optimal in the short term. It is desirable to have algorithms which are long-term optimal. To achieve this, the problem of computing the minimum point of an unconstrained function is formulated as a sequence of optimal control problems. Some qualitative results are obtained from the optimal control analysis. These qualitative results are then used to construct a theoretical iterative method and a new continuous-time method for computing the minimum point of a nonlinear unconstrained function. New iterative algorithms which approximate the theoretical iterative method and the proposed continuous-time method are then established. For convergence analysis, it is useful to note that the numerical solution of an unconstrained optimization problem is none other than an inverse Lyapunov function problem. Convergence conditions for the proposed continuous-time method and iterative algorithms are established by using the Lyapunov function theorem.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 40
    Electronic Resource
    Electronic Resource
    Springer
    Periodica mathematica Hungarica 14 (1983), S. 259-267 
    ISSN: 1588-2829
    Keywords: Primary 10A35 ; Secondary 10F40, 10K05 ; Second order linear recurrence ; fractional part ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 41
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 35 (1981), S. 45-55 
    ISSN: 1573-2878
    Keywords: Epsilon method ; numerical solutions ; relaxed controls ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract A relaxed control problem is solved using the epsilon method of Balakrishnan. The singularity of the problem with respect to one of the controls presents no difficulty. However, solutions are sensitive to the initial guess of the states and controls. For a system with true dynamics, terminal conditions may generally be approximated, but cannot be satisfied exactly.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 42
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 88 (1996), S. 709-724 
    ISSN: 1573-2878
    Keywords: Multipliers ; penalty functions ; duality ; nonsmooth optimization ; stochastic subgradients ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract This paper presents a stochastic algorithm with proper stopping rules for nonsmooth inequality-constrained minimization problems. The algorithm is based on an augmented Lagrangian dual problem transformed from a primal one, and it consists of two loops: an outer loop, which is the iteration for the approximate Lagrange multipliers, and an inner loop, which is a nonsmooth unconstrained minimization subroutine. Under mild assumptions, the algorithm is proved to be almost surely convergent.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 43
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 96 (1998), S. 575-588 
    ISSN: 1573-2878
    Keywords: Global optimization ; convergence ; stochastic algorithms ; deterministic algorithms
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract There are many global optimization algorithms which do not use global information. We broaden previous results, showing limitations on such algorithms, even if allowed to run forever. We show that deterministic algorithms must sample a dense set to find the global optimum value and can never be guaranteed to converge only to global optimizers. Further, analogous results show that introducing a stochastic element does not overcome these limitations. An example is simulated annealing in practice. Our results show that there are functions for which the probability of success is arbitrarily small.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 44
    Electronic Resource
    Electronic Resource
    Springer
    Journal of global optimization 10 (1997), S. 257-281 
    ISSN: 1573-2916
    Keywords: Global optimization ; multiextremal algorithms ; Lipschitzian first derivatives ; convergence ; numerical experiments.
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In this paper we propose a new multi-dimensional methodto solve unconstrained global optimization problems with Lipschitzianfirst derivatives. The method is based on apartition scheme that subdivides the search domain into a set of hypercubesin the course of optimization. This partitioning is regulated by thedecision rule that provides evaluation of the "importance"of each generated hypercube and selection of some partition element to performthe next iteration. Sufficient conditions of global convergence for the newmethod are investigated. Results of numerical experiments are alsopresented.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 45
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 103 (1999), S. 603-622 
    ISSN: 1573-2878
    Keywords: Monotone variational inequalities ; decomposition methods ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In the solution of the monotone variational inequality problem VI(Ω, F), with $$u = \left[ {\begin{array}{*{20}c} x \\ y \\ \end{array} } \right],Fu = \left[ {\begin{array}{*{20}c} {fx - ATy} \\ {Ax - b} \\ \end{array} } \right],\Omega = \mathcal{X} \times \mathcal{Y},$$ the augmented Lagrangian method (a decomposition method) is advantageous and effective when $$\mathcal{X} = \mathcal{R}^m$$ . For some problems of interest, where both the constraint sets $$\mathcal{X}$$ and $$\mathcal{Y}$$ are proper subsets in $$\mathcal{R}^n$$ and $$\mathcal{R}^m$$ , the original augmented Lagrangian method is no longer applicable. For this class of variational inequality problems, we introduce a decomposition method and prove its convergence. Promising numerical results are presented, indicating the effectiveness of the proposed method.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 46
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 37 (1982), S. 315-341 
    ISSN: 1573-2878
    Keywords: Optimization ; unconstrained minimization ; updates ; line searches ; convergence ; numerical methods
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract We consider a certain generalization of the Huang family of updates and discuss, firstly, convergence, dependence on parameters, and descent property; secondly, invariance under nonlinear scaling, conjugacy of search directions, and possibility of achieving a better approximation of the inverse of the Hessian. The last three aspects are shown to be dependent on particular choices of parameters. A numerical experiment is presented comparing the performances of the usual and modified BFGS algorithms.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 47
    Electronic Resource
    Electronic Resource
    Springer
    Journal of global optimization 8 (1996), S. 189-199 
    ISSN: 1573-2916
    Keywords: Semi-infinite linear programming ; optimality ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract This paper presents a globally convergent method for solving a general semi-infinite linear programming problem. Some important features of this method include: It can solve a semi-infinite linear program having an unbounded feasible region. It requires an inexact solution to a nonlinear subproblem at each iteration. It allows unbounded index sets and nondifferentiable constraints. The amount of work at each iteration k does not increase with k.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 48
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 100 (1999), S. 389-395 
    ISSN: 1573-2878
    Keywords: Complementarity problems ; change of variables ; algorithms ; fixed points ; convergence ; Wiener–Hopf equations
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In this paper, we establish an equivalence between the generalized complementarity problems and the Wiener–Hopf equations by using a change of variables technique. This equivalence is used to suggest and analyze a number of iterative algorithms for solving the generalized complementarity problems.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 49
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 91 (1996), S. 61-79 
    ISSN: 1573-2878
    Keywords: Constrained optimization ; equality constrained problems ; penalty parameters ; nonmonotonic penalty parameters ; convergence ; trust-region methods ; first-order point ; second-order point ; necessary conditions
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In a recent paper (Ref. 1), the author proposed a trust-region algorithm for solving the problem of minimizing a nonlinear function subject to a set of equality constraints. The main feature of the algorithm is that the penalty parameter in the merit function can be decreased whenever it is warranted. He studied the behavior of the penalty parameter and proved several global and local convergence results. One of these results is that there exists a subsequence of the iterates generated by the algorithm that converges to a point that satisfies the first-order necessary conditions. In the current paper, we show that, for this algorithm, there exists a subsequence of iterates that converges to a point that satisfies both the first-order and the second-order necessary conditions.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 50
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 96 (1998), S. 667-682 
    ISSN: 1573-2878
    Keywords: Existence ; nonlinear initial-value problems ; upper and lower solutions ; convergence ; quasilinearization
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract An extension of the method of quasilinearization has been applied to first-order nonlinear initial-value problems (IVP for short). It has been shown that there exist monotone sequences which converge rapidly to the unique solution of IVP.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 51
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 99 (1998), S. 271-302 
    ISSN: 1573-2878
    Keywords: Genetic algorithms ; real-number based problems ; binary encoding ; real-number encoding ; crossover operators ; convergence
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract This paper is concerned with crossover operators for genetic algorithms (GAs) which are used to solve problems based on real numbers. First, a classification of the operators is introduced, dividing crossover into a vector-level and a variable-level operator. The theoretical study of variable-level operators for binary coded GAs leads to the discovery of two properties, which are used to define certain characteristics of crossover operators used by real-number encoded GAs. For variable-level operators, the experimental distributions of the offspring variables of given pairs of parent variables are then found. Finally, an experimental comparison of crossover operator performance is carried out.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 52
    Electronic Resource
    Electronic Resource
    Springer
    Journal of optimization theory and applications 99 (1998), S. 465-480 
    ISSN: 1573-2878
    Keywords: Existence ; periodic boundary-value problems ; upper and lower solutions ; convergence ; quasilinearization
    Source: Springer Online Journal Archives 1860-2000
    Topics: Mathematics
    Notes: Abstract In this paper, the results of Lakshmikantham et al. (Ref. 1) for first-order periodic boundary-value problems are extended, by using the extended method of quaislinearization and rapid convergence for initial-value problems of Mohapatra et al. (Ref. 2). Also, it is shown that monotone sequences converge cubically to the unique solution when the forcing function in the differential equation is 2–hyperconvex and converge quartically when the forcing function is 3–hyperconvex. Several other generalizations of the problem are also presented.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
  • 53
    Electronic Resource
    Electronic Resource
    Springer
    Statistics and computing 6 (1996), S. 201-216 
    ISSN: 1573-1375
    Keywords: Conditionally linear parameters ; convergence ; EM algorithm ; Gauss-Seidel iteration ; method of scoring ; Newton method ; non-linear estimation ; orthogonal parameters ; separable regression
    Source: Springer Online Journal Archives 1860-2000
    Topics: Computer Science , Mathematics
    Notes: Abstract There are a variety of methods in the literature which seek to make iterative estimation algorithms more manageable by breaking the iterations into a greater number of simpler or faster steps. Those algorithms which deal at each step with a proper subset of the parameters are called in this paper partitioned algorithms. Partitioned algorithms in effect replace the original estimation problem with a series of problems of lower dimension. The purpose of the paper is to characterize some of the circumstances under which this process of dimension reduction leads to significant benefits. Four types of partitioned algorithms are distinguished: reduced objective function methods, nested (partial Gauss-Seidel) iterations, zigzag (full Gauss-Seidel) iterations, and leapfrog (non-simultaneous) iterations. Emphasis is given to Newton-type methods using analytic derivatives, but a nested EM algorithm is also given. Nested Newton methods are shown to be equivalent to applying to same Newton method to the reduced objective function, and are applied to separable regression and generalized linear models. Nesting is shown generally to improve the convergence of Newton-type methods, both by improving the quadratic approximation to the log-likelihood and by improving the accuracy with which the observed information matrix can be approximated. Nesting is recommended whenever a subset of parameters is relatively easily estimated. The zigzag method is shown to produce a stable but generally slow iteration; it is fast and recommended when the parameter subsets have approximately uncorrelated estimates. The leapfrog iteration has less guaranteed properties in general, but is similar to nesting and zigzagging when the parameter subsets are orthogonal.
    Type of Medium: Electronic Resource
    Location Call Number Expected Availability
    BibTip Others were also interested in ...
Close ⊗
This website uses cookies and the analysis tool Matomo. More information can be found here...