A Note on the Hyperbolic-Secant Distribution

This note discusses briefly some basic properties of the hyperbolic-secant distribution (e.g., Johnson and Kotz 1970; Manoukian 1986; Perks 1932; Talacko 1956), which has not received sufficient attention in the published literature, and may be useful for students and practitioners.     

STEIN'S METHOD AND THE BERRY-ESSÉEN THEOREM

Stein's method is used to prove the Lindeberg-Feller theorem and a generalization of the Berry-Esséen theorem. The arguments involve only manipulation of probability inequalities, and form an attractive alternative to the less direct Fourier-analytic methods which are traditionally employed.     

Limit Theorems for the Counting Process of Near-Records

Near-records of a sequence, as defined in Balakrishnan et al. (2005 Balakrishnan , N. , Pakes , A. G. , Stepanov , A. ( 2005 ). On the number and sum of near-record observations . Advances in Applied Probability 37 : 765 – 780 .[Crossref], [Web of Science ®] , [Google Scholar]), are observations lying within a fixed distance of the current record. In this article we study the asymptotic behavior of the number of near-records, among the first n observations in a sequence of independent, identically distributed and absolutely continuous random variables. We give conditions for the finiteness of the total number of near-records as well as laws of large numbers for their counting process. For distributions with a finite number of near-records, we carry out a simulation study suggesting that the total number of near-records has a geometric distribution.     

An adaptive randomized design with application to estimation

When allocating observations to two populations for estimation or testing, the optimal proportion of the data that should be allocated to the first population, if it exists, often depends on unknown parameters. Adaptive designs have thus been proposed, in which allocation of the next observation is based on an estimate of the optimal proportion computed from the data already gathered. The authors introduce a simple randomized adaptive design and give some of its properties. Applications are given to estimating the difference of two success probabilities, and the difference of two normal means.     

Applications of quadrivariate exponential distribution to a three-unit warm standby system with dependent structure

Two-unit warm standby systems have been elaborately dealt within the literature. However, the study of standby systems with more than two units, though very relevant in state-of-the-art practical situations, has received little attention because of mathematical intricacies involved in analyzing them. Also, such systems have been studied assuming: (i) the lifetime or repair time of the units to be exponential, or (ii) the lifetime and repair time to be independent. The present contribution is an improvement in the state-of-the-art in the sense that three-unit warm standby system with dependent structure is shown to be capable of comprehensive analysis.     

On the ratio of two independent skewnesses

In many fields, the researchers are interested in making inferences about the ratio of skewnesses in two independent populations. In the present paper, the asymptotic distribution for the ratio of the sample skewnesses in two independent populations is established. Then the asymptotic distribution is used to derive the asymptotic confidence interval and to test the hypothesis for the ratio of population's skewnesses. Finally, the applicability of the proposed method is investigated through Monte Carlo simulations.     

AN APPROACH TO THE SAMPLING THEOREM FOR CONTINUOUS TIME PROCESSES

In this paper we extend the sampling theorem to the full spectrum of weakly stationary stochastic processes. The mean square approximation error bound is also given.     

A Category Representation Paradox

Simple between-group comparisons of category representation percentages can be misleading for investigation of differential group selection procedures—the presence of a strong selection factor can obscure the effects of intermediate factors. This paradoxical effect is easily resolved through elementary application of the Bayes theorem. Because category representation percentages are often convenient and intuitively appealing to the statistically naive, however, there exist diverse and pervasive opportunities for incorrect interpretation with fallacious conclusions. Toxic shock syndrome epidemiological data are used to illustrate the paradox.     

Asymptotic Properties of Random Weighted Empirical Distribution Function

This article studies the asymptotic properties of the random weighted empirical distribution function of independent random variables. Suppose X₁, X₂, ???, X_n is a sequence of independent random variables, and this sequence is not required to be identically distributed. Denote the empirical distribution function of the sequence by F_n(x). Based on the random weighting method and F_n(x), the random weighted empirical distribution function H_n(x) is constructed and the asymptotic properties of H_n are discussed. Under weak conditions, the Glivenko–Cantelli theorem and the central limit theorem for the random weighted empirical distribution function are obtained. The obtained results have also been applied to study the distribution functions of random errors of multiple sensors.     

Equating Poisson and Normal Probability Functions to Derive Stirling's Formula

Often texts for introductory courses in probability or mathematical statistics make reference to Stirling's asymptotic formula (for a factorial) without presenting any proof or justification for the formula. A notable exception is Feller. In this article, we present a derivation of Stirling's formula based on the normal approximation of a Poisson probability that is considerably more accessible to the average student than Feller's approach. Besides illustrating a usage of the central limit theorem in conjunction with a continuity correction, the derivation lends itself to a mnemonic device for quickly obtaining Stirling's formula.     

Boosting with Bayesian stumps

Boosting is a new, powerful method for classification. It is an iterative procedure which successively classifies a weighted version of the sample, and then reweights this sample dependent on how successful the classification was. In this paper we review some of the commonly used methods for performing boosting and show how they can be fit into a Bayesian setup at each iteration of the algorithm. We demonstrate how this formulation gives rise to a new splitting criterion when using a domain-partitioning classification method such as a decision tree. Further we can improve the predictive performance of simple decision trees, known as stumps, by using a posterior weighted average of them to classify at each step of the algorithm, rather than just a single stump. The main advantage of this approach is to reduce the number of boosting iterations required to produce a good classifier with only a minimal increase in the computational complexity of the algorithm.     

A class of time-reversible semi-Markov processes

In this paper, we investigate a class of semi-Markov processes, first touched upon lightly in Pyke and Schaufele (1966), that are time-reversible. Results that are difficult to prove for semi-Markov processes in general are often made more simple for time-reversible semi-Markov processes, and some examples are given.     

Testing the difference between two independent regression models

ABSTRACT

In some situations, for example, in biology or psychology studies, we wish to determine whether the linear relationship between response variable and predictor variables differs in two populations. The analysis of the covariance (ANCOVA) or, equivalently, the partial F-test approaches are the commonly used methods. In this study, the asymptotic distribution for the difference between two independent regression coefficients was established. The proposed method was used to derive the asymptotic confidence set for the difference between coefficients and hypothesis testing for the equality of the two regression models. Then a simulation study was conducted to compare the proposed method with the partial F method. The performance of the new method was comparable with that of the partial F method.     

Some results following from conditional characteristic functions

ABSTRACT

The purpose of this article is to derive rigorously some results following from conditional characteristic functions and anticipate that they will prove to be of significant applicability. Specifically, Khintchine weak law of large numbers, Lévy central limit theorem, inversion theorem, uniqueness theorem, characterization of identical distribution, and characterization of independence are all generalized to conditional setup.     

A central limit theorem for generalized discounting

A central limit theorem is derived for the discounted sum of a sequence of i.i.d. random variables with finite first and second moment, where the rate of interest may depend on time. For the special case of constant rate of interest GERBER (1971) obtains a Berry-Esséen result assuming finite third moment. The same result is shown to hold in the non-const ant case.     

Universal methods for generating random variables with a given characteristic function

Universal generators for absolutely-continuous and integer-valued random variables are introduced. The proposal is based on a generalization of the rejection technique proposed by Devroye [The computer generation of random variables with a given characteristic function. Computers and Mathematics with Applications. 1981;7:547–552]. The method involves a dominating function solely requiring the evaluation of integrals which depend on the characteristic function of the underlying random variable. The proposal gives rise to simple algorithms which may be implemented in a few code lines and which may show noticeable performance even if some classical families of distributions are considered.     

Generalized definition of the geometric mean of a non negative random variable

The first probabilistic definition of the geometric mean of a non negative random variable under certain assumptions was given in Feng et al. (2013 Feng, C., Wang, H., Tu, X. (2013). Geometric mean of nonnegative random variable. Commun. Stat.—Theory Methods 42:2714–2717.[Taylor &; Francis Online], [Web of Science ®] , [Google Scholar]). In this paper, we generalize the definition to a larger class of random variables. We also show the basic properties of the geometric mean and point out its discontinuity and instability. Some convergence properties are studied as well, for which we emphasize its link to the positive moments of the random variable. A discussion of potential applications of the new definition in biomedical research and open questions to complete the theory of geometric mean is highlighted.