Dual representation of φ-divergences and applications

In this Note, we give a “dual” representation of divergences. We make use of this representation to define and study some new estimates of the law and of the divergences for discrete and continuous parametric models. To cite this article: A. Keziou, C. R. Acad. Sci. Paris, Ser. I 336 (2003).     

A representation formula for transition probability densities of diffusions and applications

We establish a representation formula for the transition probability density of a diffusion perturbed by a vector field, which takes a form of Cameron–Martin's formula for pinned diffusions. As an application, by carefully estimating the mixed moments of a Gaussian process, we deduce explicit, strong lower and upper estimates for the transition probability function of Brownian motion with drift of linear growth.     

A fuzzy representation of data for relational databases

A structure for representing inexact information in the form of a relational database is presented. The structure differs from ordinary relational databases in two important respects: Components of tuples need not be single values and a similarity relation is required for each domain set of the database. Two critical properties possessed by ordinary relational databases are proven to exist in the fuzzy relational structure. These properties are (1) no two tuples have identical interpretations, and (2) each relational operation has a unique result.     

Hypergeometrics and the cost structure of quadtrees

Several characteristic parameters of randomly grown quadtrees of any dimension are analyzed. Additive parameters have expectations whose generating functions are expressible in terms of generalized hypergeometric functions. A complex asymptotic process based on singularity analysis and integral representations akin to Mellin transforms leads to explicit values for various structure constants related to path length, retrieval costs, and storage occupation.     

An integral representation for the weighted geometric mean and its applications

By virtue of Cauchy’s integral formula in the theory of complex functions,the authors establish an integral representation for the weighted geometric mean,apply this newly established integral representation to show that the weighted geometric mean is a complete Bernstein function,and find a new proof of the well-known weighted arithmetic-geometric mean inequality.     

On determinantal representation for the generalized inverse A and its applications

In this paper, first we establish a determinantal representation for the group inverse A_g of a square matrix A. Based on this, a determinantal representation for the generalized inverse A is presented. As an application, we give a determinantal formula for the unique solution of the general restricted linear system: Ax=b(x ∈ T, b ∈ AT and dim(AT)=dim(T)), which reduces to the common Cramer rule if A is non‐singular. These results extend our earlier work. Copyright © 2006 John Wiley & Sons, Ltd.     

An efficient representation of Benes networks and its applications

The most popular bounded-degree derivative network of the hypercube is the butterfly network. The Benes network consists of back-to-back butterflies. There exist a number of topological representations that are used to describe butterfly—like architectures. We identify a new topological representation of butterfly and Benes networks.The minimum metric dimension problem is to find a minimum set of vertices of a graph G(V,E) such that for every pair of vertices u and v of G, there exists a vertex w with the condition that the length of a shortest path from u to w is different from the length of a shortest path from v to w. It is NP-hard in the general sense. We show that it remains NP-hard for bipartite graphs. The algorithmic complexity status of this NP-hard problem is not known for butterfly and Benes networks, which are subclasses of bipartite graphs. By using the proposed new representations, we solve the minimum metric dimension problem for butterfly and Benes networks. The minimum metric dimension problem is important in areas such as robot navigation in space applications.     

Mixture representation for order statistics from INID progressive censoring and its applications

In this paper, a mixture representation for the joint distribution function of progressively Type-II censored order statistics from heterogeneous distributions is established. Applications of this representation to stochastic orderings and inequalities are then illustrated.     

A new representation for second order stochastic integral-differential operators and its applications

In this paper, we’ll prove new representation theorems for a kind of second order stochastic integraldifferential operator by stochastic differential equations(SDEs) and backward stochastic differential equations(BSDEs) with jumps, and give some applications.     

Sparse time-frequency representation of nonlinear and nonstationary data

Adaptive data analysis provides an important tool in extracting hidden physical information from multiscale data that arise from various applications. In this paper, we review two data-driven time-frequency analysis methods that we introduced recently to study trend and instantaneous frequency of nonlinear and nonstationary data. These methods are inspired by the empirical mode decomposition method (EMD) and the recently developed compressed (compressive) sensing theory. The main idea is to look for the sparsest representation of multiscale data within the largest possible dictionary consisting of intrinsic mode functions of the form {a(t) cos(θ(t))}, where a is assumed to be less oscillatory than cos(θ(t)) and θ′ ? 0. This problem can be formulated as a nonlinear l ⁰ optimization problem. We have proposed two methods to solve this nonlinear optimization problem. The first one is based on nonlinear basis pursuit and the second one is based on nonlinear matching pursuit. Convergence analysis has been carried out for the nonlinear matching pursuit method. Some numerical experiments are given to demonstrate the effectiveness of the proposed methods.     

Dyadic representation of the Rudin-Shapiro coefficients with applications

The coefficients of the Rudin-Shapiro polynomials are ±1. In this paper we first replace—1 coefficient by 0 which on that case the structure of the coefficients will be on base 2. Then using the results obtained for the numbers on base 2, we introduce a quite fast algorithm to calculate the autocorrelation coefficients. Main facts: Regardless of frequencies, finding the autocorrelations of those polynomials on which their coefficients lie in the unit disk has been a telecommunication's demand. The Rudin-Shapiro polynomials have a very special form of coefficients that allow us to use “Machine language” for evaluating these values.     

Martingale representation for Poisson processes with applications to minimal variance hedging

We consider a Poisson process η on a measurable space equipped with a strict partial ordering, assumed to be total almost everywhere with respect to the intensity measure λ of η. We give a Clark-Ocone type formula providing an explicit representation of square integrable martingales (defined with respect to the natural filtration associated with η), which was previously known only in the special case, when λ is the product of Lebesgue measure on R₊ and a σ-finite measure on another space X. Our proof is new and based on only a few basic properties of Poisson processes and stochastic integrals. We also consider the more general case of an independent random measure in the sense of Itô of pure jump type and show that the Clark-Ocone type representation leads to an explicit version of the Kunita-Watanabe decomposition of square integrable martingales. We also find the explicit minimal variance hedge in a quite general financial market driven by an independent random measure.     

A q-integral representation of Rogers' q-ultraspherical polynomials and some applications

Aq-integral representation of Rogers'q-ultraspherical polynomialsC _n (x;β∥q) is obtained by using Sears' summation formula for balanced non-terminating₃ φ ₂ series. It is then used to give a simple derivation of the Gasper-Rahman formula for the Poisson kernel ofC _n (x;β∥q). As another application it is shown how this representation can be directly used to give an asymptotic expansion of theq-ultraspherical polynomials.     

An efficient random forests algorithm for high dimensional data classification

In this paper, we propose a new random forest (RF) algorithm to deal with high dimensional data for classification using subspace feature sampling method and feature value searching. The new subspace sampling method maintains the diversity and randomness of the forest and enables one to generate trees with a lower prediction error. A greedy technique is used to handle cardinal categorical features for efficient node splitting when building decision trees in the forest. This allows trees to handle very high cardinality meanwhile reducing computational time in building the RF model. Extensive experiments on high dimensional real data sets including standard machine learning data sets and image data sets have been conducted. The results demonstrated that the proposed approach for learning RFs significantly reduced prediction errors and outperformed most existing RFs when dealing with high-dimensional data.     

DB-HReduction: A data preprocessing algorithm for data mining applications

Data preprocessing is an important and critical step in the data mining process and it has a huge impact on the success of a data mining project. In this paper, we present an algorithm DB-HReduction, which discretizes or eliminates numeric attributes and generalizes or eliminates symbolic attributes very efficiently and effectively. This algorithm greatly decreases the number of attributes and tuples of the data set and improves the accuracy and decreases the running time of the data mining algorithms in the later stage.     

Pyramids: A data type for matrix representation in Pascal

Pyramids, a new data structure for the representation of matrices, are introduced. The motivation for this is Strassen's algorithm for matrix multiplication. The basic operations for pyramids are described in Pascal.Work carried out under National Research Council of Canada Grants Nos. A8327 and A7700.     

Properties of a representation of a basis for the null space

Given a rectangular matrixA(x) that depends on the independent variablesx, many constrained optimization methods involve computations withZ(x), a matrix whose columns form a basis for the null space ofA ^T(x). WhenA is evaluated at a given point, it is well known that a suitableZ (satisfyingA ^T Z = 0) can be obtained from standard matrix factorizations. However, Coleman and Sorensen have recently shown that standard orthogonal factorization methods may produce orthogonal bases that do not vary continuously withx; they also suggest several techniques for adapting these schemes so as to ensure continuity ofZ in the neighborhood of a given point.This paper is an extension of an earlier note that defines the procedure for computingZ. Here, we first describe howZ can be obtained byupdating an explicit QR factorization with Householder transformations. The properties of this representation ofZ with respect to perturbations inA are discussed, including explicit bounds on the change inZ. We then introduceregularized Householder transformations, and show that their use implies continuity of the full matrixQ. The convergence ofZ andQ under appropriate assumptions is then proved. Finally, we indicate why the chosen form ofZ is convenient in certain methods for nonlinearly constrained optimization.The research of the Stanford authors was supported by the U.S. Department of Energy Contract DE-AM03-76SF00326, PA No. DE-AT03-76ER72018; the National Science Foundation Grants MCS-7926009 and ECS-8312142; the Office of Naval Research Contract N00014-75-C-0267; and the U.S. Army Research Office Contract DAAG29-84-K-0156.The research of G.W. Stewart was supported by the Air Force Office of Scientific Research Contract AFOSR-82-0078.