Change point tests based on U-statistics with applications in reliability

We consider the problem of testing the null hypothesis of no change against the alternative of exactly one change point. The proposed tests are based on generalized two-sample U-statistic processes. We drive the limiting null distributions of the proposed tests. Some applications in Statistical Reliability are given. This research was supported by an NSERC Canada grant at the University of Alberta. Part of this research was done while visiting the University of Alberta supported by the NSERC Canada grant of the first author.     

On testing for lorenz ordering

Summary A nonparametric asymptotically distribution free approach for testing for Lorenz ordering is presented. Both the one-sample case and the two sample case are considered. This research was supported by an NSERC Canada operating grant at the University of Alberta and Research grant SM065 of Kuwait University.     

On testing exponentiality against IFRA alternatives

Summary We present a class of tests for exponentiality against IFRA alternatives. The class of tests of Deshpande (1983) is a subclass of ours. We also treat the same problem when the data is randomly censored from the right. The results of an asymptotic relative efficiency comparison indicate the superiority of our tests. This research was supported by an NSERC Canada operating grant at the University of Alberta.     

Inference ony>y o by using a correlated variable: A three decision approach

Summary In an extension of the two decision approach [Bauer, Scheiber andWohlzogen, 1975] a Bayes solution is aimed at for the three decisiony>y _o,yy _o or no classification on the basic of the measurement of a positively correlated random variableX, which can be measured more easily and/or with smaller expense. Assuming a bivariate normal distribution forX andY optimal decision regions for the measuredx are derived in the case of constant or exponentially increasing losses.

---

Zusammenfassung In Erweiterung des Zwei-Entscheidungsproblems [Bauer, Scheiber undWohlzogen, 1975] wird eine Bayes-Lösung für die drei Entscheidungeny>y ₀,yy ₀ oder keine Zuordnung aufgrund der Messung einer mitY positiv korrelierten, einfacher und/oder billiger zugänglichen ZufallsvariablenX angestrebt. Optimale Entscheidungsbereiche für die Messungenx werden bei Voraussetzung einer bivariaten Normalverteilung fürX undY unter der Annahme konstanter oder exponentiell wachsender Verluste bestimmt.

     

Two-sample rank tests with truncated populations

LetX ₁,…,X _m andY ₁,…,Y _n be two independent samples from continuous distributionsF andG respectively. Using a Hoeffding (1951) type theorem, we obtain the distributions of the vector S=(S ₍₁₎,…,S _(n)), whereS _(j)=# (X _i ’s≤Y _(j)) andY _(j) is thej-th order statistic ofY sample, under three truncation models: (a)G is a left truncation ofF orG is a right truncation ofF, (b)F is a right truncation ofH andG is a left truncation ofH, whereH is some continuous distribution function, (c)G is a two tail truncation ofF. Exploiting the relation between S and the vectorR of the ranks of the order statistics of theY-sample in the pooled sample, we can obtain exact distributions of many rank tests. We use these to compare powers of the Hajek test (Hajek 1967), the Sidak Vondracek test (1957) and the Mann-Whitney-Wilcoxon test. We derive some order relations between the values of the probagility-functions under each model. Hence find that the tests based onS ₍₁₎ andS _(n) are the UMP rank tests for the alternative (a). We also find LMP rank tests under the alternatives (b) and (c).     

Zur Entscheidung y>y0 aufgrund der Messung einer korrelierten Variablen

Zusammenfassung Es wird eine optimale Strategie im Sinne des minimalen erwarteten Verlustes für die beiden Entscheidungeny>y _o undyy _o aufgrund der Messungen einer mitY positiv korrelierten, einfacher und/oder billiger zugänglichen ZufallsvariablenX abgeleitet. Dabei wird angenommen, daßX undY nach einer bivariaten Normalverteilung mit bekannten Parametern verteilt sind und die Entscheidungyy _o getroffen wird, wennx größer ist als ein zu bestimmendesx _o, und die Entscheidungy>y _o, wennx gleich oder kleiner als diesesx _o ist. Für die Bestimmung des optimalenx _o werden zunächst die Kosten für die beiden Fehlentscheidungen jeweils als konstant vorausgesetzt, in einem weiteren Ansatz wird jedoch für die Mißklassifikationyy _o eine mity exponentiell wachsende Risikofunktion angenommen. Um die relative Häufigkeit der zu erwartenden Fehlklassifikationen abschätzen zu können, wird schließlich die bedingte WahrscheinlichkeitP(x>x _o,y) errechnet.

---

Summary An optimal strategy, with minimum expected risk, for the decisionsy>y _o oryy _o is constructed on the basis of the measurement of a variableX, which is positively correlated withY and can be measured more easily and/or with smaller expense. A bivariate normal distribution with known parameters is assumed forX andY. For the observationsx a limitx _o is aimed at, so that the decisionsy>y _o oryy _o are taken ifx>x orxx _o respectively. Optimal values ofx _o are first calculated under the assumption of constant losses for the two misclassifications (x>x _o ifyy _o andxx _o ify>y _o). In a further approach the loss for a wrong decisionyy _o is assumed to increase exponentially withy. Finally the conditional probabilityP (x>x _o\y) is calculated to get an assessment of the relative frequencies of wrong decisions to be expected.

     

Estimation of P[Y < X] for generalized exponential distribution

This paper deals with the estimation of P[Y < X] when X and Y are two independent generalized exponential distributions with different shape parameters but having the same scale parameters. The maximum likelihood estimator and its asymptotic distribution is obtained. The asymptotic distribution is used to construct an asymptotic confidence interval of P[Y < X]. Assuming that the common scale parameter is known, the maximum likelihood estimator, uniformly minimum variance unbiased estimator and Bayes estimator of P[Y < X] are obtained. Different confidence intervals are proposed. Monte Carlo simulations are performed to compare the different proposed methods. Analysis of a simulated data set has also been presented for illustrative purposes.Part of the work was supported by a grant from the Natural Sciences and Engineering Research Council     

On the estimation ofPr{Y<X} for the two-parameter exponential distribution

In this paper the Blackwell-Rao and Lehmann-Scheffé theorems are used to derive the minimum variance unbiased estimator ofP=Pr{Y when the independent random variablesX andY follow the two-parameter exponential distribution. Following a Bayesian approach, an estimator ofP is also obtained for this distribution. These results are extended for the case of censored samples.     

Some remarks on stochastic dominance

The present work proposes a definition of dominance (dominance in the strict sense), which is weaker than first order stochastic dominance, stating precisely that the r.v.Y dominatesX (XY) if Pr(YX)=1.Such a dominance in the strict sense is then compared with first and second order stochastic dominance and with dominance between descisions of the same decision problem summarised in a table of results, arriving at certain general remarks about decision problems and the choice between r.v.'s. Indications are also given about how it is possible to obtain simple and useful bounds for Pr(YX).

---

Riassunto Nel presente lavoro si propone una definizione di dominanza (dominanza in senso stretto) più debole della dominanza stocastica del prim'ordine, precisamente dicendo che la v.a.Y dominaX (XY) se Pr(YX)=1.Si confronta poi tale dominanza in senso stretto con le dominanze stocastiche del primo e del secondo ordine e con la dominanza tra decisioni di uno stesso problema di decisione sintetizzato in una tabella dei risultati giungendo ad alcune precisazioni generali sui problemi di decisione e di scelta tra v.a. Si danno anche indicazioni su come sia possibile ottenere limitazioni per la Pr(YX).

     

Sulla scelta tra variabili aleatorie

Date due variabili aleatorie stocasticamente indipendentiX eY può accadere cheX dominiY secondo il criterio della dominanza stocastica FSD anche seP(X>Y)0. In questo lavoro l'autore propone un nuovo criterio di dominanzaH fondato sulla teoria dell'utilità SSB e lo applica al caso di due variabili aleatorie dipendenti.X può dominareY secondo il criterioH solo seP(X>Y)0,5. Nel caso di variabili aleatorie indipendenti il criterioH risulta essere un affinamento del criterio FSD.

---

Summary Given two independent random variablesX andY it can happen thatX dominatesY according to the usual stochastic dominance criterion FSD even thoughP(X>Y)0. In this paper, the author proposes a new criterionH involving the SSB utility theory (Fishburn 1982) and applies it to the case of dependent random variables. It happens thatX can dominateY according to the criterionH only ifP(X>Y)0,5.In the case of independent variables, the criterionH is finer than the usual one, i.e., ifX dominatesY according to FSD then the same is true with respect toH.

     

Prediction of <Emphasis Type="Italic">k</Emphasis>-records from a general class of distributions under balanced type loss functions

We study the problem of predicting future k-records based on k-record data for a large class of distributions, which includes several well-known distributions such as: Exponential, Weibull (one parameter), Pareto, Burr type XII, among others. With both Bayesian and non-Bayesian approaches being investigated here, we pay more attention to Bayesian predictors under balanced type loss functions as introduced by Jafari Jozani et al. (Stat Probab Lett 76:773–780, 2006a). The results are presented under the balanced versions of some well-known loss functions, namely squared error loss, Varian’s linear-exponential loss and absolute error loss or L ₁ loss functions. Some of the previous results in the literatures such as Ahmadi et al. (Commun Stat Theory Methods 34:795–805, 2005), and Raqab et al. (Statistics 41:105–108, 2007) can be achieved as special cases of our results. Partial support from Ordered and Spatial Data Center of Excellence of Ferdowsi University of Mashhad is acknowledged by J. Ahmadi. M. J. Jozani’s research supported partially by a grant of Statistical Research and Training Center. é. Marchand’s research supported by NSERC of Canada. A. Parsian’s research supported by a grant of the Research Council of the University of Tehran.     

Tendency towards normality of linear combinations of random variables

IfX andY are two random variables with the same means and variances, thenX is said to be nearer normal thanY if the absolute values of its cumulants are smaller than the corresponding cumulants ofY. Using this definition, it is shown that a linear combination of a finite number of independent identically distributed random variables is always nearer normal than its constituents, but that this is not necessarily true if not-identically distributed or not-independent variables are used. Some consequences of the results are reached for the testing of normality of time series and for the assumptions frequently made by social scientists about the distribution of their data.     

Convergence of a branching type recursion with non-stationary immigration

The asymptotic distribution of a branching type recursion with non-stationary immigration is investigated. The recursion is given by , where (X _l ) is a random sequence, (L _n ⁻¹⁽¹⁾ ) are iid copies ofL _n−1,K is a random number andK, (L _n ⁻¹⁽¹⁾ ), {(X _l ),Y _n } are independent. This recursion has been studied intensively in the literature in the case thatX _l ≥0,K is nonrandom andY _n =0 ∀n. Cramer, Rüschendorf (1996b) treat the above recursion without immigration with starting conditionL ₀=1, and easy to handle cases of the recursion with stationary immigration (i.e. the distribution ofY _n does not depend on the timen). In this paper a general limit theorem will be deduced under natural conditions including square-integrability of the involved random variables. The treatment of the recursion is based on a contraction method. The conditions of the limit theorem are built upon the knowledge of the first two moments ofL _n . In case of stationary immigration a detailed analysis of the first two moments ofL _n leads one to consider 15 different cases. These cases are illustrated graphically and provide a straight forward means to check the conditions and to determine the operator whose unique fixed point is the limit distribution of the normalizedL _n .     

Almost unbiased estimation in multiplicative models

In the lognormal linear models the estimation of constant term presents problems. In this paper we use weighted jackknife procedure (suggested by Hinkley 1977) for reducing the bias of the maximum likelihood estimator. The resulting estimator is unbiased upto order (1/T),T being the number of observations, and has the same MSE as that of the MLE to the same order of approximation; moreover, being the jackknife estimator it enjoys all the desirable large sample properties like any other jackknife estimator. The research of this author is partially supported through a research grant from NSERC of Canada.     

Über die Anzahl von Zufallspunkten mit typ-gleichem nächsten Nachbarn und einen multivariaten Zwei-Stichproben-Test

Summary For independents-variate samplesX ₁, ...,X _m i.i.d.f. (.),Y ₁, ...,Y _n i.i.d. g. (.), where the densitiesf (.),g (.) are assumed to be continuous on their respective sets of positivity, consider the numberT _m,n of pointsZ of the pooled sample (which are either of typeX or of typeY) such that the nearest neighbor ofZ is of the same type asZ. We show that, as , independently of (.). An omnibus test for the two sample problem f(.)g(.) orf(.)g(.)? may be obtained by rejecting the hypothesisf(.)g(.) for large values ofT _m,n.     

A consistent characteristic function-based test for conditional independence

Y is conditionally independent of Z given X   if Pr{f(y|X,Z)=f(y|X)}=1$\{f(y|X,Z)=f(y|X)\}=1$ for all y on its support, where f(·|·)$f(·|·)$ denotes the conditional density of Y   given (X,Z)$(X,Z)$ or X.$X.$ This paper proposes a nonparametric test of conditional independence based on the notion that two conditional distributions are equal if and only if the corresponding conditional characteristic functions are equal. We extend the test of Su and White (2005. A Hellinger-metric nonparametric test for conditional independence. Discussion Paper, Department of Economics, UCSD) in two directions: (1) our test is less sensitive to the choice of bandwidth sequences; (2) our test has power against deviations on the full support of the density of (X,Y,Z$X,Y,Z$). We establish asymptotic normality for our test statistic under weak data dependence conditions. Simulation results suggest that the test is well behaved in finite samples. Applications to stock market data indicate that our test can reveal some interesting nonlinear dependence that a traditional linear Granger causality test fails to detect.     

A measure of mutual complete dependence

Two random variables X and Y on a common probability space are mutually completely dependent (m.c.d.) if each one is a function of the other with probability one. For continuous X and Y, a natural approach to constructing a measure of dependence is via the distance between the copula of X and Y and the independence copula. We show that this approach depends crucially on the choice of the distance function. For example, the L ^p -distances, suggested by Schweizer and Wolff, cannot generate a measure of (mutual complete) dependence, since every copula is the uniform limit of copulas linking m.c.d. variables. Instead, we propose to use a modified Sobolev norm, with respect to which mutual complete dependence cannot approximate any other kind of dependence. This Sobolev norm yields the first nonparametric measure of dependence which, among other things, captures precisely the two extremes of dependence, i.e., it equals 0 if and only if X and Y are independent, and 1 if and only if X and Y are m.c.d. Examples are given to illustrate the difference to the Schweizer–Wolff measure.