On a class of functions with equal infima over a domain and its boundary

In this paper, we present a class of functions:f:X such that inf _xX f(x)= , whereX is a nonempty, finitely compact and convex set in a vector space andB _x ={xX: y aff(X){x:[x, y]X={x}. Our main tool is a recent minimax theorem by Ricceri (Ref. 1).     

Self-bounded controlled invariant subspaces: A straightforward approach to constrained controllability

A lattice-type structure is shown to exist in a particular subset of the set of all (A, )-controlled invariants contained in and containing , whereA denotes a linear map inR ⁿ ; , are arbitrary subspaces ofR ⁿ ; andD is an arbitrary subspace ofJ, the maximum (A, )-controlled invariant contained in . In linear system theory, this property can be used for a more direct theoretical and algorithmic approach to constrained controllability and disturbance rejection problems.     

The Maximum Box Problem and its Application to Data Analysis

Given two finite sets of points X ⁺ and X ^– in ⁿ , the maximum box problem consists of finding an interval (box) B = {x : l x u} such that B X ^– = , and the cardinality of B X ⁺ is maximized. A simple generalization can be obtained by instead maximizing a weighted sum of the elements of B X ⁺. While polynomial for any fixed n, the maximum box problem is -hard in general. We construct an efficient branch-and-bound algorithm for this problem and apply it to a standard problem in data analysis. We test this method on nine data sets, seven of which are drawn from the UCI standard machine learning repository.     

QUADRATIC ESTIMATORS OF QUADRATIC FUNCTIONS WITH PARAMETERS IN NORMAL LINEAR MODELS

ＱＵＡＤＲＡＴＩＣＥＳＴＩＭＡＴＯＲＳＯＦＱＵＡＤＲＡＴＩＣＦＵＮＣＴＩＯＮＳＷＩＴＨＰＡＲＡＭＥＴＥＲＳＩＮＮＯＲＭＡＬＬＩＮＥＡＲＭＯＤＥＬＳ￥ＷＵＱＩＧＵＡＮＧ（吴启光）（ＩｎｓｔｉｔｕｔｅｏｆＳｙｓｔｅｍｅＳｃｉｅｎｃｅ,ｔｈｅＣｈｉｎｅｓｅ...     

Sequential gradient-restoration algorithm for optimal control problems

This paper considers the problem of minimizing a functionalI which depends on the statex(t), the controlu(t), and the parameter . Here,I is a scalar,x ann-vector,u anm-vector, and ap-vector. At the initial point, the state is prescribed. At the final point, the statex and the parameter are required to satisfyq scalar relations. Along the interval of integration, the state, the control, and the parameter are required to satisfyn scalar differential equations. Asequential algorithm composed of the alternate succession of gradient phases and restoration phases is presented. This sequential algorithm is contructed in such a way that the differential equations and boundary conditions are satisfied at the end of each iteration, that is, at the end of a complete gradient-restoration phase; hence, the value of the functional at the end of one iteration is comparable with the value of the functional at the end of any other iteration.In thegradient phase, nominal functionsx(t),u(t), satisfying all the differential equations and boundary conditions are assumed. Variations x(t), u(t), leading to varied functions (t),(t), are determined so that the value of the functional is decreased. These variations are obtained by minimizing the first-order change of the functional subject to the linearized differential equations, the linearized boundary conditions, and a quadratic constraint on the variations of the control and the parameter.Since the constraints are satisfied only to first order during the gradient phase, the functions (t),(t), may violate the differential equations and/or the boundary conditions. This being the case, a restoration phase is needed prior to starting the next gradient phase. In thisrestoration phase, the functions (t),(t), are assumed to be the nominal functions. Variations (t), (t), leading to varied functions (t),û(t), consistent with all the differential equations and boundary conditions are determined. These variations are obtained by requiring the least-square change of the control and the parameter subject to the linearized differential equations and the linearized boundary conditions. Of course, the restoration phase must be performed iteratively until the cumulative error in the differential equations and boundary conditions becomes smaller than some preselected value.If the gradient stepsize is , an order-of-magnitude analysis shows that the gradient corrections are x=O(), u=O(), =O(), while the restoration corrections are . Hence, for sufficiently small, the restoration phase preserves the descent property of the gradient phase: the functionalI decreases between any two successive iterations.Methods to determine the gradient stepsize in an optimal fashion are discussed. Examples are presented for both the fixed-final-time case and the free-final-time case. The numerical results show the rapid convergence characteristics of the sequential gradient-restoration algorithm.The portions of this paper dealing with the fixed-final-time case were presented by the senior author at the 2nd Hawaii International Conference on System Sciences, Honolulu, Hawaii, 1969. The portions of this paper dealing with the free-final-time case were presented by the senior author at the 20th International Astronautical Congress, Mar del Plata, Argentina, 1969. This research, supported by the NASA-Manned Spacecraft Center, Grant No. NGR-44-006-089, Supplement No. 1, is a condensation of the investigations presented in Refs. 1–5. The authors are indebted to Professor H. Y. Huang for helpful discussions.     

Approximations for optimal Laplace transform inversion

A function (p) of the Laplace transform operatorp is approximated by a finite linear combination of functions (p+ _r ), where (p) is a specific function ofp having a known analytic inverse (t), and is chosen in accordance with various considerations. Then parameters _r ,r=1, 2,...,n, and then corresponding coefficientsA _r of the (p + _r ) are determined by a least-square procedure. Then, the corresponding approximation to the inversef(t) of (p) is given by analytic inversion of _r=1 ⁿ A _r (p+ _r ). The method represents a generalization of a method of best rational function approximation due to the author [which corresponds to the particular choice (t)1], but is capable of yielding considerably greater accuracy for givenn.The computations for this paper were carried out on the CDC-6600 computer at the Computation Center of Tel-Aviv University. The author is grateful to Dr. H. Jarosch of the Weizmann Institute of Science Computer Center for use of their Powell minimization subroutine (Ref. 1).     

Steepest Descent with Curvature Dynamical System

Let H be a real Hilbert space and let <..,.> denote the corresponding scalar product. Given a function that is bounded from below, we consider the following dynamical system:

Torsion in K‐Theory for Boundary Actions on Affine Buildings of Type ?n

Let be a torsionfree lattice in G=PGL(n+1, , where n 1 and is a nonArchimedean local field. Then acts on the Furstenberg boundary G/P, where P is a minimal parabolic subgroup of G. The identity element I in the crossedproduct C ^*algebra C(G/P) generates a class [I] in the K ₀ group of C(G/P) . It is shown that [I] is a torsion element of K ₀ and there is an explicit bound for the order of [I]. The result is proved more generally for groups acting on affine buildings of type à _n. For n=1, 2 the Euler–Poincaré characteristic () annihilates the class [I].     

Inadmissibility and admissibility results for unbiased loss estimators based on gauss-markov estimators

LetY be distributed according to ann-variate normal distribution with a meanX and a nonsingular covariance matrix ² V, where bothX andV are known, R ^p is a parameter, > 0 is known or unknown. Denote and . Assume thatF is linearly estimable. When is known, it is proved that the unbiased loss estimator ²tr(F(XV ^–1 X)^– F) of is admissible for rank (F)=k4 and inadmissible fork 5 with the squared error loss . When is unknown and rank (X) <n, it is established that the loss estimatorcS ², wherec is any nonnegative constant, of is inadmissible and that the unbiased loss estimator tr(F(XV ^–1 X)^– F) of is admissible fork 4, and inadmissible fork 5 with squared error loss.This project is supported by the National Natural Science Foundation of China.     

Cotype of operators fromC(K)

Summary Forq>2, an operator fromC(K) toX is of cotypeq if and only if it factors through the Lorentz space . Forq=2, ifX is a rearrangement invariant space on [0, 1], the injectionC([0, 1])X is of cotype 2 if and only if it factors through the Lorentz space ; but there is a cotype 2 operator C(K) that does not factor through . If a Banach latticeX satisfies the Orlicz property, any bounded lattice operatorT:C(K)X is of cotype 2. We however construct a Banach lattice with the Orlicz property, but that fails to be of cotype 2.Oblatum 4-VII-1990 & 18-IV-1991Work partially supported by an NSF grant     

On the Point Spectrum of Self-Adjoint Operators That Appears under Singular Perturbations of Finite Rank

We discuss purely singular finite-rank perturbations of a self-adjoint operator A in a Hilbert space . The perturbed operators are defined by the Krein resolvent formula , Im z 0, where B _z are finite-rank operators such that dom B _z dom A = |0}. For an arbitrary system of orthonormal vectors satisfying the condition span | _i } dom A = |0} and an arbitrary collection of real numbers , we construct an operator that solves the eigenvalue problem . We prove the uniqueness of under the condition that rank B _z = n.     

A variant of an infinite-dimensional local limit theorem

Let _n be a sequence of independent, identically distributed random elements in a separable Banach space X, for which the CLTholds: the normalized sums (₁+...+_n)/n^1/2 converge weakly to the Gaussian random element . It is proved that, under certain conditions on the distribution of ₁ and on the measurable mappingf: X R¹, the distribution of the random variable converges in variation to the distribution of the variablef().Translated from Zapiski Nauchnykh Seminarov Leningradskogo Otdeleniya Matematicheskogo Instituta im. V. A. Steklova Akademii Nauk SSSR, Vol. 177, pp. 46–50, 1989.     

Differential Equations in Hilbert Space with Dissipative Symbols

In this article equations of the form

Scalar-quadratic stabilizability of the Petersen counterexample via a linear static controller

A new property called scalar-quadratic is presented for establishing the stabilizability of linear time-varyring uncertain systems. It is applied to a well-known linear time-varying system _OL which contains two uncertainties ₁(t) and ₂(t). Using the Lyapunov functionsV(x)=x ^T Px, whereP is a constant postitive-definite symmetric matrix, previous authors have shown that _OL is stabilizable by linear static controllers when the time-varying uncertainties are bounded by a normalized bound satisfying < 0.8. We extend the bound to < 1.0 by using the more general Lyapunov functions satisfying the scalar-quadratic propertyV(ax)=a ² V(x), aR, xR ₀ ² .Our proof uses a hexagon as a closed, convex hypersuface to construct a scalar-quadratic Lyapunov function, so that the Lyapunov time derivative satisfies the quadratic convergence condition , >0, for the closed-loop system _CL formed from _OL and a stabilizing linear static controller. The critical condition in the proof of the quaratic convergence ondition is the satisfaction of the inequality , where _max is a normalization bound for ₁(t) and ₂(t) and wheree ₁ ande ₂ are parameters for the controller. For the controller parametrized bye ₁=8 ande ₂=20, this inequality reduces to _max < 2.2096. This result, in particular, establishes that the Petersen counterexample is stabilitzable by the linear static controller withe ₁=8 ande ₂=20. Furthermore, it establishes the amazing result that _OL is stabilizable by a linear static controlle on any compact subset of the constant uncertainaty controllability space defined by ₁>0 and ₂>0.     

Hilbert Series of Group Representations and Gröbner Bases for Generic Modules

Each matrix representation :G GL_n() of a finite Group G over a field induces an action of G on the module ⁿ over the polynomial algebra The graded -submodule M() of _n generated by the orbit of is studied. A decomposition of M() into generic modules is given. Relations between the numerical invariants of and those of M(), the latter being efficiently computable by Gröbner bases methods, are examined. It is shown that if is multiplicity-free, then the dimensions of the irreducible constituents of can be read off from the Hilbert series of M(Pi;). It is proved that determinantal relations form Gröbner bases for the syzygies on generic matrices with respect to any lexicographic order. Gröbner bases for generic modules are also constructed, and their Hilbert series are derived. Consequently, the Hilbert series of M(Pi;) is obtained for an arbitrary representation.     

Large Deviation Probabilities for Some Classes of Distributions Satisfying the Cramer Condition

In the present paper, we study large deviations of the sum X₁ + + X_n of i.i.d. random variables under the assumption that < for some > 0. Bibliography: 17 titles.__________Translated from Zapiski Nauchnykh Seminarov POMI, Vol. 298, 2003, pp. 161–185.     

Matroid Shellability, β-Systems,and Affine Hyperplane Arrangements

The broken-circuit complex is fundamental to the shellability and homology of matroids, geometric lattices, and linear hyperplane arrangements. This paper introduces and studies the -system of a matroid, nbc(M), whose cardinality is Crapo's -invariant. In studying the shellability and homology of base-pointed matroids, geometric semilattices, and afflne hyperplane arrangements, it is found that the -system acts as the afflne counterpart to the broken-circuit complex. In particular, it is shown that the -system indexes the homology facets for the lexicographic shelling of the reduced broken-circuit complex , and the basic cycles are explicitly constructed. Similarly, an EL-shelling for the geometric semilattice associated with M is produced,_and it is shown that the -system labels its decreasing chains.Basic cycles can be carried over from The intersection poset of any (real or complex) afflnehyperplane arrangement is a geometric semilattice. Thus the construction yields a set of basic cycles, indexed by nbc(M), for the union of such an arrangement.     

Theorems of the Alternative and Duality

This paper investigates the relations between theorems of the alternative and the minimum norm duality theorem. A typical theorem of the alternative is associated with two systems of linear inequalities and/or equalities, a primal system and a dual one, asserting that either the primal system has a solution, or the dual system has a solution, but never both. On the other hand, the minimum norm duality theorem says that the minimum distance from a given point z to a convex set is equal to the maximum of the distances from z to the hyperplanes separating z and . We consider the theorems of Farkas, Gale, Gordan, and Motzkin, as well as new theorems that characterize the optimality conditions of discrete l ₁-approximation problems and multifacility location problems. It is shown that, with proper choices of , each of these theorems can be recast as a pair of dual problems: a primal steepest descent problem that resembles the original primal system, and a dual least–norm problem that resembles the original dual system. The norm that defines the least-norm problem is the dual norm with respect to that which defines the steepest descent problem. Moreover, let y solve the least norm problem and let r denote the corresponding residual vector. If r=0, which means that z , then y solves the dual system. Otherwise, when r0 and z , any dual vector of r solves both the steepest descent problem and the primal system. In other words, let x solve the steepest descent problem; then, r and x are aligned. These results hold for any norm on . If the norm is smooth and strictly convex, then there are explicit rules for retrieving x from r and vice versa.     

Pluriadjoint bundles of polarized surfaces

LetX be a complex connected projective smooth algebraic surface and letL be an ample line bundle onX. The maps associated with the pluriadjoint bundles (K _X L) ¹,t2, are studied by combining an ampleness result forK _X L with a very recent result by Reider. It turns out that apart from some exceptions and up to reductions, 1) (K _X L)³ is very ample; 2) (K _X L) ² is ample and spanned by global sections and is very ample unless eitherg (L)=2 (arithmetic genus ofL) orX contains an elliptic curveE withE ²=0,E·L=1;3) when (K _X L) ² is not very ample, the associated map has degree 4, equality implying thatg (L)=2 and .     

A dissipative transformation with a trivial tail-algebra

Summary In [1], an example was given of a measure-preserving dissipative transformation T in a -finite measure space (X, , ), such that T is conservative in the measure space (X, , ) where . Here we shall show that for this transformation we actually have R ={ØX}[].