Subdifferentiability of Convex Functions with Values in an Ordered Vector Space.

Random variables can be described by their cumulative distribution functions, a class of nondecreasing functions on the real line. Those functions can in turn be identified, after the possible vertical gaps in their graphs are filled in, with maximal monotone relations. Such relations are known to be the subdifferentials of convex functions. Analysis of these connections yields new insights. The generalized inversion operation between distribution functions and quantile functions corresponds to graphical inversion of monotone relations. In subdifferential terms, it corresponds to passing to conjugate convex functions under the Legendre---Fenchel transform. Among other things, this shows that convergence in distribution for sequences of random variables is equivalent to graphical convergence of the monotone relations and epigraphical convergence of the associated convex functions. Measures of risk that employ quantiles (VaR) and superquantiles (CVaR), either individually or in mixtures, are illuminated in this way. Formulas for their calculation are seen from a perspective that reveals how they were discovered. The approach leads further to developments in which the superquantiles for a given distribution are interpreted as the quantiles for an overlying "superdistribution." In this way a generalization of Koenker---Basset error is derived which lays a foundation for superquantile regression as a higher-order extension of quantile regression.

§1. Introduction.The motivation of this work comes from two different directions: infinite abelian groups, and ordered algebraic structures. A challenging problem in both cases is that of classification. In the first case, it is known for example (cf. [KA]) that the classification of abelian torsion groups amounts to that of reducedp-groups by numerical invariants called theUlm invariants(given by Ulm in [U]). Ulm's theorem was later generalized by P. Hill to the class of totally projective groups. As to the second case, let us consider for instance the class of divisible ordered abelian groups. These may be viewed as ordered ℚ-vector spaces. Their theory being unstable, we cannot hope to classify them by numerical invariants. On the other hand, being o-minimal, the theory enjoys several good model theoretic properties (cf. [P-S]), so the search for some reasonable invariants is well motivated. The common denominator of the two cases, as well as of many others, is valuation theory. Indeed given an ordered vector space, one can consider it as a valued vector space, endowed with the natural valuation. Also, the socleG[p] of a reduced abelianp-groupG, endowed with the height functionhG, is a valued vector space over(the prime field of characteristicp) with values in the ordinals (cf. [F]).

Fuzzy topological ordered vector spaces I

The crucial discovery of the concept of differential calculus is due to Pierre de Fermat (1601–1655), who was one of the most important innovators in the history of mathematics. It is to him that we owe a rule for determining extrema, described, without proof, in a short treatise Methodus ad disquirendam Maximum et Minimam written in 1637. The importance of his discoveries in number theory has eclipsed the contributions which this exceptional and modest man made to other areas of mathematics. Fermat also was the first to discover the “principle of least time” in optics, the prototype of the variational principles governing so many physical and mechanical laws. He shared independently with Descartes the invention of analytic geometry and with Pascal the creation of the mathematical theory of probability. His achievements in number theory overshadowed his other contributions, as the Last Fermat Theorem which remained a challenge for such a long time, and still is a challenge if indeed the simple proof of Piere de Fermat did exist. Not to mention his compositions in French, Latin, Italian and Spanish verse and his Grecian erudition. It is also notable that he was able to find time for these occupations in the midst of his duties as counsellor to the parliament of Toulouse (even taking into account Fermat’s genius, this makes us reflect on the leisure activities offered by a lawyer’s career).

In this chapter, controlled differential inclusions in a Hilbert space containing subdifferentials of convex functions are considered. The following three problems are studied: the problem of prescribed motion realization, the problem of robust control, and the problem of input dynamical reconstruction. Solution algorithms that are stable with respect to informational noises and computational errors are presented. The algorithms are based on the method of feedback control. They adaptively take into account inaccurate measurements of phase trajectories and are regularized in the following sense: the more precise is incoming information, the better is the algorithm’s output.

Subdifferentials of convex functions

Variational formulations of non-linear and non-smooth structural problems

In a partly ordered space the orthogonality relation is defined by incomparability. We define integrally open and integrally semi-open ordered real vector spaces. We prove: if an ordered real vector space is integrally semi-open, then a complete lattice of double orthoclosed sets is orthomodular. An integrally open concept is closely related to an open set in the Euclidean topology in a finite dimensional ordered vector space. We prove: if V is an ordered Euclidean space, then V is integrally open and directed (and is also Archimedean) if and only if its positive cone, without vertex 0, is an open set in the Euclidean topology (and also the family of all order segments \( \{ z \in V:a < z < b\} \), a < b, is a base for the Euclidean topology).

This paper proposes a modified Gram-Schmidt algorithm for dimension reduction of a document vector space in order to do classification. We also evaluate the performance of the proposed algorithm by comparing with commonly known algorithms such as the centroid-based algorithm and latent semantic indexing. Further, to measure efficiency, a modified Gram-Schmidt training set is applied in the centroid-based algorithm and latent semantic indexing as well. Performance measurement was based on two different parameters, viz., classification accuracy and closeness of similarity with the corresponding training set. The results shows that modified Gram-Schmidt algorithm is indeed an effective method for dimension reduction prior classification. Moreover, it is easy to code and computationally inexpensive.KeywordsCentroidDocument ClassificationDimension ReductionGram-SchmidtLSIVector Space Model

Learning bilingual sentiment lexicon for online reviews

If the vector space \({\mathcal{L}}^r(E, F)\) of all regular operators between the vector lattices E and F is ordered by the collection of its positive operators, then the Dedekind completeness of F is a sufficient condition for \({\mathcal{L}}^r(E, F)\) to be a vector lattice. \({\mathcal{L}}^r(E, F)\) and some of its subspaces might be vector lattices also in a more general situation. In the paper we deal with ordered vector spaces \({\mathcal{V}}(E, F)\) of linear operators and ask under which conditions are they vector lattices, lattice-subspaces of the ordered vector space \({\mathcal{L}}^r(E, F)\) or, in the case that \({\mathcal{L}}^r(E, F)\) is a vector lattice, sublattices or even Banach lattices when equipped with the regular norm. The answer is affirmative for many classes of operators such as compact, weakly compact, regular AM-compact, regular Dunford-Pettis operators and others if acting between appropriate Banach lattices. Then it is possible to study the finite elements in such vector lattices \({\mathcal{V}}(E, F)\), where F is not necessary Dedekind complete. In the last part of the paper there will be considered the question how the order structures of E, F and \({\mathcal{V}}(E, F)\) are mutually related. It is also shown that those rank one and finite rank operators, which are constructed by means of finite elements from E′ and F, are finite elements in \({\mathcal{V}}(E,F)\). The paper contains also some generalization of results obtained for the case \({\mathcal{V}}(E, F) = {\mathcal{L}}^r(E, F)\) in [10].

Linear mappings on ordered vector spaces

Riesz homomorphisms between vector lattices are characterized by interval preserving order adjoints. Buskes, van Rooij, and van Haandel generalized the notion of a Riesz homomorphism to linear maps between general ordered vector spaces. In this paper, we discuss how far the characterization between vector lattices extends to directed ordered vector spaces and pre-Riesz spaces.

A general extended functional is a functional which is allowed to take on infinite values; in other words such functionals are similar to the Lebesgue integral on the space of all integrable functions. The problem of representing a general extended functional f on a vector lattice X as an operator with finite values only has been solved in [12]. In fact, this problem has been solved for a larger class of general extended operators on an ordered vector space. The solution of this problem was given by means of an extension of the range R of the functional f to some ultrapower of R. Notice, however, that it is not always the case that a functional f can be considered as a trace of some internal functional *f:*X→R. (We remark, without going into details, that such an internal functional exists exactly in the case when the results of the nonstandard analysis can be used for investigation of the given functional.) In [12] a ‘standard’ necessary and sufficient condition was given for solving this latter problem on the existence of *f. Namely, f is a trace of an internal *f if and only if it is finitely extendable. This result makes the finite extendability problem worthy of study. The present paper is a first attempt in this direction. Simultaneously we introduce and study a more general notion of weak finite extendability that coincides with finite extendability, for instance, for vector lattices with a strong unit.

Book Review: Barrelledness in topological and ordered vector spaces