In this thesis the chance-constrained linear goal programming approach is developed to cover the following cases when the parameters have non-negative distributions: the exponential and the 'chi-square distributions. Case 1, when the right hand side coefficients are exponential or chi-square random vdriables. Case 2. when the input coefficients are exponential or chisquare random variables. The following have been achieved: For Case 1 1. We have developed a method for constructing deterministic linear goal programs equivalent to the original probabilistic linear goal programs. 2. We have given a probabilistic interpretation to the deviational random variables and the deviational random variable levels. For Case 2 We have developed a method for constructing deterministic nonlinear goal programs through the definition of the probabilistic deviational variables. 4. We have transformed the equivalent deterministic nonlinear goal programs into equivalent signomial goal programs. S. We have developed a computational algorithm for solving nonlinear goal programs generally and, more particularly, deterministic nonlinear goal programs equivalent to chance-constrained goal programs. iii 6. We have proved that Sengupta! s-transformation for obtaining deterministic programs equivalent to chanceconstrained programs does-not lead to solvable programs. 7. We have-formulated and solved a practical application - namely that of finding the "optimal distribution of exports and, imports to the marine, ports" using the methods and the algorithm presented in the thesis. The methods can be used when a program has mixed goals, some with right hand side coefficients or input coefficients that are exponential or chi-square random variables; -others, deterministic, that is without random variable parameters.

No description available.

L’objectif de cette thèse est d’étendre une méthode d’estimation des paramètres d’une loi stable dans Rd aux lois à queue régulière dans un cône arbitraire. La méthode d’échantillonnage par paquets est modifiée afin d’optimiser la vitesse de convergence des estimateurs. Un nouvel estimateur de la masse totale de mesure spectrale est proposé. Des résultats sur les statistiques d’ordre des lois à queue régulière dans un cône et la loi des grands nombres pour le schéma triangulaire sont établis. La consistance et la normalité asymptotique des estimateurs sont démontrées. La performance des estimateurs est étudiée par simulation. On compare ces estimateurs avec quelques estimateurs connus. Les tableaux de performance sont fournis. La méthode de noyau est utilisée pour estimer la densité d’une mesure spectrale absolument continue d’une loi à queue régulière. On prouve la consistance de l’estimateur dans notre cas particulier. Pour augmenter le nombre de points utilisés dans l’échantillon, on propose une méthode d’estimation utilisant les permutations aléatoires de l’échantillon. La variation régulière a la propriété d’être préservée par plusieurs opérations et transformations. On considère trois sortes de transformations. Des conditions suffisantes pour cette préservation sont proposées et quelques contre-exemples sont présentés. Les modèles de lois stables et de lois à queue lourde sont très utilisés dans plusieurs domaines d’application. On considère deux jeux de données réelles : les cours des 30 valeurs de l’indice DJIA et les perturbations planétaires des comètes du nuage de Oort. En appliquant la méthode d’estimation présentée on obtient des descriptions statistiques de ces données. / The objective of this thesis is to extend an estimation method of parameters of a stable distribution in Rd to the regularly varying tail distributions in an arbitrary cone. The sampling method of regrouping is modified to optimize the rate of convergence of estimators. A new estimator of total mass of the spectral measure is proposed. Some results about order statistics of regularly varying tail laws in a cone and the strong law of large numbers on the triangular schema are established. The consistency and the asymptotic normality of estimators are proved. The performance of proposed estimators is studied by simulation. We compare these estimators with some known estimators. The performance tables are provided. The kernel density estimation is used to estimate the density of an absolutely continuous spectral measure of a regularly varying tail law. We prove the consistency of the estimator in our particular case. To increase the number of points used in the sample, an estimation method using the random permutations of sample is proposed. The property of regular variation can be preserved by several operations and transformations. We consider three kinds of transformations. The sufficient conditions for this preservation are proposed and some counter-examples are presented. The models of stable distributions and heavy tailed distributions are widely used in several application areas. We consider two sets of real data : the prices of 30 stocks of the DJIA index and the planetary perturbations of comets of the Oort cloud. By applying the estimation method presented previously we obtain some statistical descriptions of these data.

Transit agencies have the opportunity to improve the delivery of services by using data from Intelligent Transportation Systems (ITS). On-time performance is an important measure. The objective of this paper is to adjust the timetables so that the probability of on-time performance is maximized. For this purpose we analyze data distributions of travel time and also consider the general case that data distribution is unknown. Statistical procedures are presented to find scheduled time for some selected distributions. Monte Carlo simulation is introduced for the purpose of finding scheduled time when data distribution is not known. Simulation studies indicate that the on-time performance would increase using the proposed methodology. The contribution of this paper is to provide transit system a procedure to set up or update their timetables based on current ITS data and its distribution, and hence increase level of service.

No description available.

No description available.

No / We document rich variation across observed firms’ characteristics, and the accompanying macroeconomic volatility, often related to political turmoil for Ukrainian manufacturing firms. We use a unique annual firm-level data for the period from 2001 to 2009 compiled from the Derzhkomstat. To understand the evolution of distributions we utilize functional principal component analysis while accounting for the effects associated with firms’ region, industry, trade status, and firm turnover. The overall improvements in firm productivity in Ukraine’s manufacturing in 2001–2009 vary substantially by industry, trade status and with firm turnover, while regional effects are less important.

Este trabalho apresenta uma demonstração completa do Teorema de L. Hörmander sobre a divisão de distribuições (temperadas) por polinômios. O caso n=1 é apresentado detalhadamente e serve como motivação para as técnicas utilizadas no caso geral. Todos os pré-requisitos para a demonstração de Hörmander (os Teoremas de Seidenberg-Tarski, de Puiseux e da Extensão de Whitney) são discutidos com detalhes. Como conseqüência do Teorema, segue que todo operador diferencial parcial linear com coeficientes constantes não nulo admite solução fundamental temperada. / This dissertation presents a thorough proof of L. Hörmander\'s theorem on the division of (tempered) distributions by polynomials. The case n=1 is discussed in detail and serves as a motivation for the techniques that are utilised in the general case. All the prerequisites for Hörmander\'s proof (the Theorems of Seidenberg-Tarski, of Puiseux and Whitney\'s Extension Theorem) are discussed in detail. As a consequence of this theorem, it follows that every non zero partial diffe\\-rencial operator with constant coefficients has a tempered fundamental solution.

This thesis addresses some problems that arise in signal processing when the noise is impulsive and follows a heavy tailed distribution. After reviewing several of the more well known heavy- tailed distributions the common problem of which of these hest models the observations is considered. To this end, a test is proposed for the symmetric alpha stable distribution. The test threshold is found using both asymptotic theory and parametric bootstrap resampling. In doing so, some modifications are proposed for Koutrouvelis' estimator of the symmetric alpha stable distributions parameters that improve performance. In electrical systems impulsive noise is generated externally to the receiver while thermal Gaussian noise is generated internally by the receiver electronics, the resultant noise is an additive combination of these two independent sources. A characteristic function domain estimator for the parameters of the resultant distribution is developed for the case when the impulsive noise is modeled by a symmetric alpha stable distribution. Having concentrated on validation and parameter estimation for the noise model, some problems in signal detection and estimation are considered. Detection of the number of sources impinging on an array is an important first. step in many array processing problems for which the development of optimal methods can be complicated even in the Gaussian case. Here, a multiple hypothesis test for the equality of the eigenvalues of the sample array covariance is proposed. / The nonparametric bootstrap is used to estimate the distributions of the test statistics removing the assumption of Gaussianity and offering improved performance for heavy tailed observations. Finally, some robust estimators are proposed for estimating parametric signals in additive noise. These are based on M-estimators but implicitly incorporate an estimate of the noise distribution. enabling the estimator to adapt to the unknown noise distribution. Two estimators are developed, one uses a nonparametric kernel density estimator while the other models the score function of the noise distribution with a linear combination of basis functions.

Motivated by applications in multimedia streaming and in energy systems, we study duality relations in fi nite queues. Dual of a queue is de fined to be a queue in which the arrival and service processes are interchanged. In other words, dual of the G1/G2/1/K queue is the G2/G1/1/K queue, a queue in which the inter-arrival times have the same distribution as the service times of the primal queue and vice versa. Similarly, dual of a fluid flow queue with cumulative input C(t) and available processing S(t) is a fluid queue with cumulative input S(t) and available processing C(t). We are primarily interested in finding relations between the overflow and underflow of the primal and dual queues. Then, using existing results in the literature regarding the probability of loss and the stationary probability of queue being full, we can obtain estimates on the probability of starvation and the probability of the queue being empty. The probability of starvation corresponds to the probability that a queue becomes empty, i.e., the end of a busy period. We study the relations between arrival and departure Palm distributions and their relations to stationary distributions. We consider both the case of point process inputs as well as fluid inputs. We obtain inequalities between the probability of the queue being empty and the probability of the queue being full for both the time stationary and Palm distributions by interchanging arrival and service processes. In the fluid queue case, we show that there is an equality between arrival and departure distributions that leads to an equality between the probability of starvation in the primal queue and the probability of overflow in the dual queue. The techniques are based on monotonicity arguments and coupling. The usefulness of the bounds is illustrated via numerical results.