Exact probabilities of correct classifications for uncorrelated repeated measurements from elliptically contoured distributions

Euclidean distance-based classification rules are derived within a certain nonclassical linear model approach and applied to elliptically contoured samples having a density generating function g. Then a geometric measure theoretical method to evaluate exact probabilities of correct classification for multivariate uncorrelated feature vectors is developed. When doing this one has to measure suitably defined sets with certain standardized measures. The geometric key point is that the intersection percentage functions of the areas under investigation coincide with those of certain parabolic cylinder type sets. The intersection percentage functions of the latter sets can be described as threefold integrals. It turns out that these intersection percentage functions yield simultaneously geometric representation formulae for the doubly noncentral g-generalized F-distributions. Hence, we get beyond new formulae for evaluating probabilities of correct classification new geometric representation formulae for the doubly noncentral g-generalized F-distributions. A numerical study concerning several aspects of evaluating both probabilities of correct classification and values of the doubly noncentral g-generalized F-distributions demonstrates the advantageous computational properties of the present new approach. This impression will be supported by comparison with the literature.It is shown that probabilities of correct classification depend on the parameters of the underlying sample distribution through a certain well-defined set of secondary parameters. If the underlying parameters are unknown, we propose to estimate probabilities of correct classification.     

A concept of variety for regular biordered sets

We define a bivariety of regular biordered sets to be a nonempty class of regular biordered sets which is closed under taking direct products, regular bimorphic images and relatively regular biordered subsets. It is then shown that there is a complete lattice morphism mapping the complete lattice of all e-varieties of regular semigroups onto the complete lattice of all bivarieties of regular biordered sets; as a corollary we prove that there is a complete lattice morphism mapping the complete lattice of all e-varieties of E-solid regular semigroups onto the complete lattice of all varieties of solid binary algebras. Examples of bivarieties include the class of all solid regular biordered sets and the class of all local semilattices. For each setX with at least two elements, we show that a bivariety contains a free object onX if and only if it consists entirely of solid regular biordered sets or it consists entirely of local semilattices. The author gratefully acknowledges the financial support of an Australian Postgraduate Research Award.     

Generating and Enumerating Digitally Convex Sets of Trees

A set S of vertices in a graph G with vertex set V is digitally convex if for every vertex \(v \in V\), \(N[v] \subseteq N[S]\) implies \(v \in S\). We show that a vertex belongs to at most half of the digitally convex sets of a graph. Moreover, a vertex belongs to exactly half of the digitally convex sets if and only if it is simplicial. An algorithm that generates all digitally convex sets of a tree is described and sharp upper and lower bounds for the number of digitally convex sets of a tree are obtained. A closed formula for the number of digitally convex sets of a path is derived. It is shown how a binary cotree of a cograph can be used to enumerate its digitally convex sets in linear time.     

Identifying predictive hubs to condense the training set of k-nearest neighbour classifiers

The $k$ -Nearest Neighbour classifier is widely used and popular due to its inherent simplicity and the avoidance of model assumptions. Although the approach has been shown to yield a near-optimal classification performance for an infinite number of samples, a selection of the most decisive data points can improve the classification accuracy considerably in real settings with a limited number of samples. At the same time, a selection of a subset of representative training samples reduces the required amount of storage and computational resources. We devised a new approach that selects a representative training subset on the basis of an evolutionary optimization procedure. This method chooses those training samples that have a strong influence on the correct prediction of other training samples, in particular those that have uncertain labels. The performance of the algorithm is evaluated on different data sets. Additionally, we provide graphical examples of the selection procedure.     

Characterizations of strictly convex sets by the uniqueness of support points

Abstract

This short paper characterizes strictly convex sets by the uniqueness of support points (such points are called unique support points or exposed points) under appropriate assumptions. A class of so-called regular sets, for which every extreme point is a unique support point, is introduced. Closed strictly convex sets and their intersections with some other sets are shown to belong to this class. The obtained characterizations are then applied to set-valued maps and to the separation of a convex set and a strictly convex set. Under suitable assumptions, so-called set-valued maps with path property are characterized by strictly convex images of the considered set-valued map.     

On sets of uniqueness for harmonic functions in the unit circle

The resutls of this paper show that the structure of sets mentioned in the title is not trivial. For example, it is shown that there exist countalbe sets of uniqueness for logarithmic potential, i.e., closed countable subsets E of the unit circle $\mathbb{T}$ such that $$f \in C(\mathbb{T}),f|_E = 0,U^f |_E = 0 \Rightarrow f \equiv 0.$$ Here $U^f (z) = \tfrac{1}{\pi }\int\limits_0^{2\pi } {f(e^{i\theta } )\log \tfrac{1}{{\left| {z - e^{i\theta } } \right|}}d\theta } $ . On the other hand, it is shown that every countable porous closed subset of $\mathbb{T}$ is a nonuniqueness set. Bibliography: 9 titles.     

Penalty function theory for general convex programming problems

A class of penalty functions for solving convex programming problems with general constraint sets is considered. Convergence theorems for penalty methods are established by utilizing the concept of infimal convergence of a sequence of functions. It is shown that most existing penalty functions are included in our class of penalty functions.     

Reconstruction of Discrete Sets from Four Projections: Strong Decomposability

In this paper we introduce the class of strongly decomposable discrete sets and give an efficient algorithm for reconstructing discrete sets of this class from four projections. It is also shown that every Q-convex set (along the set of directions {x, y}) consisting of several components is strongly decomposable. As a consequence of strong decomposability we get that in a subclass of hv-convex discrete sets the reconstruction from four projections can be solved in polynomial time.     

Finding a complete set of solutions or proving unsolvability for certain classes of matrix polynomial equations with commuting coefficients

Two classes of matrix polynomial equations with commuting coefficients are examined. It is shown that the equations in one class have complete sets of solutions, whereas the equations in the other class are unsolvable. A method is given for finding the solution set of an equation in the former class.     

Computable Ramsey’s theorem for pairs needs infinitely many $$\Pi ^0_2$$ sets

In Ramsey’s Theorem and Recursion Theory, Theorem 4.2, Jockusch proved that for any computable k-coloring of pairs of integers, there is an infinite \(\Pi ^0_2\) homogeneous set. The proof used a countable collection of \(\Pi ^0_2\) sets as potential infinite homogeneous sets. In a remark preceding the proof, Jockusch stated without proof that it can be shown that there is no computable way to prove this result with a finite number of \(\Pi ^0_2\) sets. We provide a proof of this claim, showing that there is no computable way to take an index for an arbitrary computable coloring and produce a finite number of indices of \(\Pi ^0_2\) sets with the property that one of those sets will be homogeneous for that coloring. While proving this result, we introduce n-trains as objects with useful combinatorial properties which can be used as approximations to infinite \(\Pi ^0_2\) sets.     

Correction and strengthening of “How large are the level sets of the Takagi function?”

The purpose of this note is to correct an error in an earlier paper by the author about the level sets of the Takagi function (Monatsh Math 167:311–331, 2012), and to prove a stronger form of one of the main results of that paper.     

Further study of production possibility set and performance evaluation model in supply chain DEA

Performance evaluation is an importance issue in supply chain management. Yang et al. (Ann. Oper. Res. 38(6):195–211, 2011) defined two types of supply chain production possibility sets and proved the equivalence between them. Based on the sub-perfect CRS production possibility set, they proposed a supply chain DEA model to appraise the overall technical efficiency of supply chains. The relationship among efficiency scores of the proposed model, CCR models of system and subsystems are discussed. However, we find that the equivalence between the two types of supply chain production possibility sets is not correct. The proofs of their three theorems are all problematic. In this paper, we correct some results and give three new proofs.     

A concentration of mass property for isotropic convex bodies in high dimensions

It is shown that, for a certain subclass of istropic convex sets in ℝⁿ, the mass concentrates in a spherical shell, asymptotically for largen. This in turn shows that the inequality is close to an equality for the mentioned class of isotropic convex sets, asymptotically for largen. It also implies a ‘central limit property’ for this class. Partly supported by the DFG.     

Improved confidence sets for the mean of a multivariate normal distribution

A new class of confidence sets for the mean of a p-variate normal distribution (p3) is introduced. They are neither spheres nor ellipsoids. We show that we can construct our confidence sets so that their coverage probabilities are equal to the specified confidence coefficient. Some of them are shown to dominate the usual confidence set, a sphere centered at the observations. Numerical results are also given which show how small their volumes are.     

An Affine Representation for Transversal Geometries

Pregeometries (matroids) whose independent sets are the partial matchings of a relation (transversal pregeometries) can be canonically imbedded in a free-simplicial pregeometry (one whose points lie freely on flats spanned by a simplex). Conversely, all subgeometries of such free-simplicial pregeometries are transversal. Free-simplicial pregeometries are counted and their duals are naturally constructed and shown to be free-simplicial (showing self-dual free-simplexes corrspond to quasisymmetric relations). For more general transversal pregeometries, modular flats are characterized and transversal contractions are exemplified. Binary transversal pregeometries and their contractions (the class of binary gammoids) are shown to be the class of series-parallel networks, providing insight for further characterizations of (coordinatized) gammoids by excluded minors. Theorem. All principal transversal pregeometries and their truncations have critical exponent at most 2.     

On the index sets of Σ-subsets of the real numbers

We compute the levels of complexity in analytical and arithmetical hierarchies for the sets of the Σ-formulas defining in the hereditarily finite superstructure over the ordered field of the reals the classes of open, closed, clopen, nowhere dense, dense subsets of ? ⁿ , first category subsets in ? ⁿ as well as the sets of pairs of Σ-formulas corresponding to the relations of set equality and inclusion which are defined by them. It is also shown that the complexity of the set of the Σ-formulas defining connected sets is at least Π ₁ ¹ .     

Novel applications of bipolar fuzzy graphs to decision making problems

Zhang introduced the concept of bipolar fuzzy sets as a generalization of fuzzy sets. Bipolar fuzzy sets have shown advantages in solving decision making problems than fuzzy sets. In this research paper, we study several different types of domination, including equitable domination, k-domination and restrained domination in bipolar fuzzy graphs. We present novel applications of bipolar fuzzy graphs to decision making problems. We also present an algorithm for computing dominating number in our applications.     

Lyapunov exponents for continuous random transformations

In this paper, the concept of Lyapunov exponent is generalized to random transformations that are not necessarily differentiable. For a class of random repellers and of random hyperbolic sets obtained via small perturbations of deterministic ones respectively, the new exponents are shown to coincide with the classical ones.