An algorithm producing a standard spine of a 3-manifold presented by surgery along a link

We provide a simple algorithm which produces a (branched) standard spine of a 3-manifold presented by surgery along a framed link inS ³, giving an explicit upper bound on the complexity of the spine in terms of the complexity of a diagram of the link. As a corollary, we get an easy constructive proof of Casler’s result on the existence of a standard spine for a closed 3-manifold. We also describe an o-graph which represents the spine.     

Center manifolds for impulsive equations under nonuniform hyperbolicity

We establish the existence of smooth center manifolds under sufficiently small perturbations of an impulsive linear equation. In particular, we obtain the C¹ smoothness of the manifolds outside the jumping times. We emphasize that we consider the general case of nonautonomous equations for which the linear part has a nonuniform exponential trichotomy.     

An iterative algorithm for a kind of HJB equations

讨论一类带T-严格单调函数HJB方程的迭代算法,并证明了算法的单调收敛性.进一步地,提出了基于此迭代算法的区域分解法.     

Generic hyperbolicity for reaction diffusion equations on symmetric domains

Summary The problem of generic hyperbolicity for reaction diffusion equations with Dirichlet boundary conditions on a ball inR ⁿ is studied. It is proved that while hyperbolicity is not a generic property, radially symmetric solutions are generically hyperbolic.

---

Zusammenfassung Das Problem der generischen Hyperbolizität für Reaktion-Diffusionsgleichungen mit Dirichletschen Randbedingungen in einer Kugel imR ⁿ wird betrachtet. Es wird bewiesen, daß radialsymmetrische Lösungen generisch symmetrische sind, während Hyperbolizität nicht eine generische Eigenschaft ist.

     

An iterative substructuring algorithm for equilibrium equations

Summary The topic of iterative substructuring methods, and more generally domain decomposition methods, has been extensively studied over the past few years, and the topic is well advanced with respect to first and second order elliptic problems. However, relatively little work has been done on more general constrained least squares problems (or equivalent formulations) involving equilibrium equations such as those arising, for example, in realistic structural analysis applications. The potential is good for effective use of iterative algorithms on these problems, but such methods are still far from being competitive with direct methods in industrial codes. The purpose of this paper is to investigate an order reducing, preconditioned conjugate gradient method proposed by Barlow, Nichols and Plemmons for solving problems of this type. The relationships between this method and nullspace methods, such as the force method for structures and the dual variable method for fluids, are examined. Convergence properties are discussed in relation to recent optimality results for Varga's theory ofp-cyclic SOR. We suggest a mixed approach for solving equilibrium equations, consisting of both direct reduction in the substructures and the conjugate gradient iterative algorithm to complete the computations.Dedicated to R. S. Varga on the occasion of his 60th birthdayResearch completed while pursuing graduate studies sponsored by the Department of Mathematical Sciences, US Air Force Academy, CO, and funded by the Air Force Institute of Technology, WPAFB, OHResearch supported by the Air Force under grant no. AFOSR-88-0285 and by the National Science Foundation under grant no. DMS-89-02121     

An algorithm for solving Fredholm integro-differential equations

The aim of this paper is to present an efficient numerical procedure for solving linear second order Fredholm integro-differential equations. The scheme is based on B-spline collocation and cubature formulas. The analysis is accompanied by numerical examples. The results demonstrate reliability and efficiency of the proposed algorithm.      

Genericity of hyperbolicity and saddle-node bifurcations in reaction-diffusion equations depending on a parameter

Semilinear elliptic equations of the form $$\begin{array}{*{20}c} {\sum\limits_{i,j = 1}^n {(a_{ij} (x)u_{xi} (x))_{x_j } + } f(\lambda ,x,u(x)) = 0,} & {x \in \Omega ,} \\ {u(x) = 0,} & x \\ \end{array} $$ are considered, where λ ε ? is a parameter, Ω ? ? ⁿ is a bounded domain andf is a smooth non-linear function. It is shown that for ‘generic’ functionsf, the set of non-trivial solutions (λ,u) consists of a finite, or countable, collection of smooth, 1-dimensional curves and any such solution is either hyperbolic or is a saddle-node bifurcation point of the curve.     

Compressing thin spheres in the complement of a link

Let L be a link in S³ that is in thin position but not in bridge position and let P be a thin level sphere with compressing disk D. We introduce the idea of alternating level spheres for D and show that all such spheres are thin and their widths are monotone decreasing. This allows us to generalize a result of Wu by giving a bound on the number of disjoint irreducible compressing disks P can have in terms of the width of P, including identifying thin spheres with unique compressing disks. We also give conditions under which P must be incompressible on some side or be weakly incompressible. In particular we show that the thin level sphere of second lowest width is weakly incompressible. If P is strongly compressible we describe how a pair of compressing disks must lie relative to the link.     

An iterative aggregation-disaggregation algorithm for solving linear equations

A general procedure is given for solving large sets of linear equations by first rewriting them in a form suitable for aggregation of both the variables and equations, followed by disaggregation. A computational algorithm which iteratively aggregates and disaggregates is shown to converge geometrically to the exact solution. Provided the original problem has a structure suitable for such aggregation, the algorithm exhibits fast computation times, small main-memory requirements, and robustness to the starting point. A rigorous foundation for aggregation and disaggregation is provided by the equations employed by this algorithm.     

An algorithm for solving linear Volterra integro-differential equations

An efficient numerical procedure for solving linear second order Volterra integro-differential equations is presented herein. The scheme is based on B-spline collocation and cubature formulas. Analysis is accompanied by numerical examples. Results confirm reliability and efficiency of the proposed algorithm.     

An improved trust region algorithm for nonlinear equations

In this paper, we present a new trust region algorithm for the system of singular nonlinear equations with the regularized trust region subproblem. The new algorithm preserves the global convergence of the traditional trust region algorithm, and has the quadratic convergence under some suitable conditions. Finally, some numerical results are given.     

Plenty of hyperbolicity on a class of linear homogeneous jerk differential equations

Aequationes mathematicae - We consider $$3times 3$$ partially hyperbolic linear differential systems over an ergodic flow $$X^t$$ and derived from the linear homogeneous differential equation...     

An ABS algorithm for a class of systems of stochastic linear equations

This paper is to explore a model of the ABS algorithms dealing with the solution of a class of systems of linear stochastic equations Aξ=η when η is a m-dimensional normal distribution. It is shown that the stepsize α _i is distributed as N(u _i ,σ _i ) (being u _i the expected value of α _i and σ _i its variance) and the approximation to the solutions ξ _i is distributed as N _n (U _i ,Σ _i ) (being U _i the expected value of ξ _i and Σ _i its variance), for this algorithm model.     

Sampling from the complement of a polyhedron: An MCMC algorithm for data augmentation

We present an MCMC algorithm for sampling from the complement of a polyhedron. Our approach is based on the Shake-and-bake algorithm for sampling from the boundary of a set and provably covers the complement. We use this algorithm for data augmentation in a machine learning task of classifying a hidden feasible set in a data-driven optimization pipeline. Numerical results on simulated and MIPLIB instances demonstrate that our algorithm, along with a supervised learning technique, outperforms conventional unsupervised baselines.     

An Arnoldi based algorithm for large algebraic Riccati equations

In the present work, we present a numerical method for the computation of approximate solutions for large continuous-time algebraic Riccati equations. The proposed method is a method of projection onto a matrix Krylov subspace. We use a matrix Arnoldi process to construct an orthonormal basis. We give some theoretical results and numerical experiments for large problems.     

极大似然估计算法研究

将解一元方程的二分法推广至求解多元非线性方程组.以第K个变元Xk为参数,则κ元方程组就可以看作曲线s(前κ-1个方程)和κ-1维曲面C(第κ个方程),于是κ元方程组的解就可以看作寻找曲线s和曲面C的交点.对参数Xk作二分法,重复迭代,直到找到满足误差要求的方程组的解.最后给出了用多元二分法的算法求解极大似然估计的数值解.