The analysis of survival data with a possible cure fraction has attracted much interest in the last two decades. Various models and estimating methods have been proposed for such data and they have been applied in many fields, especially in cancer clinical trials. In the thesis, we consider some new general cure models, which include existing survival models as their special cases. We also consider a nonparametric estimation of cure rate. The estimator is proved consistent and asymptotically normal. We also consider the application of proportional density for cure data and the analysis of length-biased cure data. / Thesis (Ph.D, Mathematics & Statistics) -- Queen's University, 2012-01-26 09:53:08.127

This thesis is about some issues in system modeling: The first is a parsimonious representation of MISO Hammerstein system, which is by projecting the multivariate linear function into a univariate input function space. This leads to the so-called semiparamtric Hammerstein model, which overcomes the commonly known “Curse of dimensionality” for nonparametric estimation on MISO systems. The second issue discussed in this thesis is orthogonal expansion analysis on a univariate Hammerstein model and hypothesis testing for the structure of the nonlinear subsystem. The generalization of this technique can be used to test the validity for parametric assumptions of the nonlinear function in Hammersteim models. It can also be applied to approximate a general nonlinear function by a certain class of parametric function in the Hammerstein models. These techniques can also be extended to other block-oriented systems, e.g, Wiener systems, with slight modification. The third issue in this thesis is applying machine learning and system modeling techniques to transient stability studies in power engineering. The simultaneous variable section and estimation lead to a substantially reduced complexity and yet possesses a stronger prediction power than techniques known in the power engineering literature so far.

This thesis is about some issues in system modeling: The first is a parsimonious representation of MISO Hammerstein system, which is by projecting the multivariate linear function into a univariate input function space. This leads to the so-called semiparamtric Hammerstein model, which overcomes the commonly known “Curse of dimensionality” for nonparametric estimation on MISO systems. The second issue discussed in this thesis is orthogonal expansion analysis on a univariate Hammerstein model and hypothesis testing for the structure of the nonlinear subsystem. The generalization of this technique can be used to test the validity for parametric assumptions of the nonlinear function in Hammersteim models. It can also be applied to approximate a general nonlinear function by a certain class of parametric function in the Hammerstein models. These techniques can also be extended to other block-oriented systems, e.g, Wiener systems, with slight modification. The third issue in this thesis is applying machine learning and system modeling techniques to transient stability studies in power engineering. The simultaneous variable section and estimation lead to a substantially reduced complexity and yet possesses a stronger prediction power than techniques known in the power engineering literature so far.

This dissertation, which consists of three essays, studies three applications. Each of them emphasizes the microfoundations of economic models. The first essay proposes a nonparametric estimation of structural labor supply and exact welfare change under nonconvex piecewise-linear budget sets. Different from previous literature, my method focuses on a nonparametric specification of an indirect utility function. I find that working with the indirect utility function is very useful in simultaneously addressing the labor supply problems with individual heterogeneity, nonconvex budget sets, labor nonparticipation, and measurement errors in working hours that previous literature was unable to. Further, the estimated indirect utility function proves to be convenient and efficient in calculating exact welfare change and deadweight loss under general piecewise-linear budget sets. In the second essay, I solve the equity premium, risk-free rate, and capital structure puzzles by laying a more solid microfoundation for consumption-based asset pricing models. I argue that the above two asset pricing puzzles arise from the aggregation of hump-shaped life-cycle consumption into per capita consumption, which accounts for the unanimous rejections of Euler equations in the literature. As for the third puzzle, I show that a firm's capital structure can be determined by heterogenous investors maximizing life-time utility even though the capital structure is irrelevant on the firm side. The endogenously determined leverage generates an even larger equity premium than a fixed one. The third essay studies the solution concepts of coalition equilibrium. Traditional solution concepts such as Strong Nash Equilibrium, Coalition-proof Nash Equilibrium, Largest Consistent Set, and Coalition Equilibrium violate the fundamental principles of individual rationality. I define a new solution concept, Weak Coalition Equilibrium, which requires each coalitional deviation to be within-coalition self-enforceable and cross-coalition self-enforceable. The cross-coalition self-enforceability endows coalitions with farsightedness. Weak Coalition Equilibrium is a generalization of Coalition-proof Nash Equilibrium and a re nement of the concept Nash Equilibrium. It exists under a weak condition. Most importantly, it is in line with the principle of individual rationality.

The underlying aim of this research is to investigate the mathematical descriptions of homogeneous textures in digital images for the purpose of segmentation and recognition. The research covers the problem of testing these mathematical descriptions by using them to generate synthetic realisations of the homogeneous texture for subjective and analytical comparisons with the source texture from which they were derived. The application of this research is in analysing satellite or airborne images of the Earth's surface. In particular, Synthetic Aperture Radar (SAR) images often exhibit regions of homogeneous texture, which if segmented, could facilitate terrain classification. In this thesis we present noncausal, nonparametric, multiscale, Markov random field (MRF) models for recognising and synthesising texture. The models have the ability to capture the characteristics of, and to synthesise, a wide variety of textures, varying from the highly structured to the stochastic. For texture synthesis, we introduce our own novel multiscale approach incorporating a new concept of local annealing. This allows us to use large neighbourhood systems to model complex natural textures with high order statistical characteristics. The new multiscale texture synthesis algorithm also produces synthetic textures with few, if any, phase discontinuities. The power of our modelling technique is evident in that only a small source image is required to synthesise representative examples of the source texture, even when the texture contains long-range characteristics. We also show how the high-dimensional model of the texture may be modelled with lower dimensional statistics without compromising the integrity of the representation. We then show how these models -- which are able to capture most of the unique characteristics of a texture -- can be for the ``open-ended'' problem of recognising textures embedded in a scene containing previously unseen textures. Whilst this technique was developed for the practical application of recognising different terrain types from Synthetic Aperture Radar (SAR) images, it has applications in other image processing tasks requiring texture recognition.

Diferente das medidas de associação global (coeficiente de correlação linear de Pearson, de Spearman, tau de Kendall, por exemplo), as medidas de dependência local descrevem o comportamento da dependência localmente em diferentes regiões. Nesta tese, as medidas de dependência local para variáveis aleatórias propostas por Bairamov et al. (2003), Bjerve e Doksum (1993) e Sibuya (1960), são estudadas sob o enfoque de processos estocásticos estacionários bivariados e univariados, neste caso, estudando o comportamento da dependência local ao longo das defasagens da série temporal. Para as duas primeiras medidas, discutimos as suas propriedades, e estudamos os seus estimadores, além da consistência dos mesmos. Para a medida de Sibuya, além de discutir suas propriedades, propomos três estimadores para variáveis aleatórias e dois para séries temporais, verificando a consistência dos mesmos. O comportamento das três medidas locais e dos seus estimadores foram avaliados através de simulações e aplicações a dados reais (neste caso, fizemos uma comparação destas com cópula e densidade cópula). / Unlike global association measures (Pearson´s linear correlation coefficient, Spearman´s rho, Kendall´s tau, for example), local dependence measures describe the behaviour of dependence locally in different regions. In this thesis, the local dependence measures for random variables proposed by Bairamov et al. (2003), Bjerve and Doksum (1993) and Sibuya (1960), are studied in the context of bivariate and univariate stationary stochastic processes, in this case, evaluating the performance of local dependence along time lags. We discussed the properties and studied the estimators and consistence of the first two measures. As for the Sibuya measure, in addition to discussing its properties, we propose three estimators for random variables and two for time series while checking their consistence. The behaviour of the three local measures and their respective estimators was evaluated by simulations and application to real data (in this case, a comparison was drawn with copula and copula density).

Cópulas tem se tornado uma importante ferramenta para descrever e analisar a estrutura de dependência entre variáveis aleatórias e processos estocásticos. Recentemente, surgiram alguns métodos de estimação não paramétricos, utilizando kernels e ondaletas. Neste contexto, sabendo que cópulas podem ser escritas como expansão em ondaletas, foi proposto um estimador não paramétrico via ondaletas para a função cópula para dados independentes e de séries temporais, considerando processos alfa-mixing. Este estimador tem como característica principal estimar diretamente a função cópula, sem fazer suposição alguma sobre a distribuição dos dados e sem ajustes prévios de modelos ARMA - GARCH, como é feito em ajuste paramétrico para cópulas. Foram calculadas taxas de convergência para o estimador proposto em ambos os casos, mostrando sua consistência. Foram feitos também alguns estudos de simulação, além de aplicações a dados reais. / Copulas are important tools for describing the dependence structure between random variables and stochastic processes. Recently some nonparametric estimation procedures have appeared, using kernels and wavelets. In this context, knowing that a copula function can be expanded in a wavelet basis, we have proposed a nonparametric copula estimation procedure through wavelets for independent data and times series under alpha-mixing condition. The main feature of this estimator is the copula function estimation without assumptions about the data distribution and without ARMA - GARCH modeling, like in parametric copula estimation. Convergence rates for the estimator were computed, showing the estimator consistency. Some simulation studies were made, as well as analysis of real data sets.

Les applications les plus courantes des méthodes non paramétriques concernent l’estimation d’une fonction de régression (i.e. de l’espérance conditionnelle). Cependant, il est souvent intéressant de modéliser les quantiles conditionnels, en particulier lorsque la moyenne conditionnelle ne permet pas de représenter convenablement l’impact des covariables sur la variable dépendante. De plus, ils permettent d’obtenir des graphiques plus compréhensibles de la distribution conditionnelle de la variable dépendante que ceux obtenus avec la moyenne conditionnelle. À l’origine, la « quantification » était utilisée en ingénierie du signal et de l’information. Elle permet de discrétiser un signal continu en un nombre fini de quantifieurs. En mathématique, le problème de la quantification optimale consiste à trouver la meilleure approximation d’une distribution continue d’une variable aléatoire par une loi discrète avec un nombre fixé de quantifieurs. Initialement utilisée pour des signaux univariés, la méthode a été étendue au cadre multivarié et est devenue un outil pour résoudre certains problèmes en probabilités numériques. Le but de cette thèse est d’appliquer la quantification optimale en norme Lp à l’estimation des quantiles conditionnels. Différents cas sont abordés : covariable uni- ou multidimensionnelle, variable dépendante uni- ou multivariée. La convergence des estimateurs proposés est étudiée d’un point de vue théorique. Ces estimateurs ont été implémentés et un package R, nommé QuantifQuantile, a été développé. Leur comportement numérique est évalué sur des simulations et des données réelles. / One of the most common applications of nonparametric techniques has been the estimation of a regression function (i.e. a conditional mean). However it is often of interest to model conditional quantiles, particularly when it is felt that the conditional mean is not representative of the impact of the covariates on the dependent variable. Moreover, the quantile regression function provides a much more comprehensive picture of the conditional distribution of a dependent variable than the conditional mean function. Originally, the “quantization” was used in signal and information theories since the fifties. Quantization was devoted to the discretization of a continuous signal by a finite number of “quantizers”. In mathematics, the problem of optimal quantization is to find the best approximation of the continuous distribution of a random variable by a discrete law with a fixed number of charged points. Firstly used for a one-dimensional signal, the method has then been developed in the multi-dimensional case and extensively used as a tool to solve problems arising in numerical probability. The goal of this thesis is to study how to apply optimal quantization in Lp-norm to conditional quantile estimation. Various cases are studied: one-dimensional or multidimensional covariate, univariate or multivariate dependent variable. The convergence of the proposed estimators is studied from a theoretical point of view. The proposed estimators were implemented and a R package, called QuantifQuantile, was developed. Numerical behavior of the estimators is evaluated through simulation studies and real data applications.

Dans cette thèse, on considère la modélisation de la loi jointe des statistiques d'ordre, c.à.d. des vecteurs aléatoires avec des composantes ordonnées presque sûrement. La première partie est dédiée à la modélisation probabiliste des statistiques d'ordre d'entropie maximale à marginales fixées. Les marginales étant fixées, la caractérisation de la loi jointe revient à considérer la copule associée. Dans le Chapitre 2, on présente un résultat auxiliaire sur les copules d'entropie maximale à diagonale fixée. Une condition nécessaire et suffisante est donnée pour l'existence d'une telle copule, ainsi qu'une formule explicite de sa densité et de son entropie. La solution du problème de maximisation d'entropie pour les statistiques d'ordre à marginales fixées est présentée dans le Chapitre 3. On donne des formules explicites pour sa copule et sa densité jointe. On applique le modèle obtenu pour modéliser des paramètres physiques dans le Chapitre 4.Dans la deuxième partie de la thèse, on étudie le problème d'estimation non-paramétrique des densités d'entropie maximale des statistiques d'ordre en distance de Kullback-Leibler. Le chapitre 5 décrit une méthode d'agrégation pour des densités de probabilité et des densités spectrales, basée sur une combinaison convexe de ses logarithmes, et montre des bornes optimales non-asymptotiques en déviation. Dans le Chapitre 6, on propose une méthode adaptative issue d'un modèle exponentiel log-additif pour estimer les densités considérées, et on démontre qu'elle atteint les vitesses connues minimax. L'application de cette méthode pour estimer des dimensions des défauts est présentée dans le Chapitre 7 / In this thesis we consider the modelling of the joint distribution of order statistics, i.e. random vectors with almost surely ordered components. The first part is dedicated to the probabilistic modelling of order statistics of maximal entropy with marginal constraints. Given the marginal constraints, the characterization of the joint distribution can be given by the associated copula. Chapter 2 presents an auxiliary result giving the maximum entropy copula with a fixed diagonal section. We give a necessary and sufficient condition for its existence, and derive an explicit formula for its density and entropy. Chapter 3 provides the solution for the maximum entropy problem for order statistics with marginal constraints by identifying the copula of the maximum entropy distribution. We give explicit formulas for the copula and the joint density. An application for modelling physical parameters is given in Chapter 4.In the second part of the thesis, we consider the problem of nonparametric estimation of maximum entropy densities of order statistics in Kullback-Leibler distance. Chapter 5 presents an aggregation method for probability density and spectral density estimation, based on the convex combination of the logarithms of these functions, and gives non-asymptotic bounds on the aggregation rate. In Chapter 6, we propose an adaptive estimation method based on a log-additive exponential model to estimate maximum entropy densities of order statistics which achieves the known minimax convergence rates. The method is applied to estimating flaw dimensions in Chapter 7

L'objectif de cette thèse est d'explorer plusieurs approches pour l'étude des données censurées multivariées, à savoir l'estimation non paramétrique de la fonction de répartition jointe, la modélisation de dépendance par les modèles de copules et l'étude exploratoire par des méthodes de clustering. Le Chapitre 1 introduit le contexte général de cette thèse ainsi que ses contributions. Le Chapitre 2 est consacré à l'estimation de la distribution jointe des deux variables censurées dans le cadre d'un modèle de durée simplifié où la différence entre deux variables de censure est observée. Un nouvel estimateur non paramétrique de la fonction de répartition jointe y est introduit. La normalité asymptotique a été démontrée, pour les intégrales par rapport à la mesure définie par cet estimateur. Le Chapitre 3 est dédié à la problématique de l'estimation non paramétrique de la copule bivariée, à partir d'un échantillon de données censurées. La copule est d'abord estimée par une fonction discrète qui peut être interprétée comme une extension de la copule empirique en présence de censure, puis par ses versions lisses. Les propriétés asymptotiques et des applications de des estimateurs ont été considérées. Le Chapitre 4 présente une approche exploratoire pour l'étude de données censurées. Plus précisément, une configuration multivariée est considérée où une variable est une durée sujette à la censure, et toutes les autres variables sont observées. Sous ces conditions, une nouvelle méthode de quantification de la loi jointe est introduite. La méthode est étudiée théoriquement et appliquée à la construction d'un algorithme de clustering pour des observations censurées. / The main purpose of this thesis is to explore several approaches for studying multivariate censored data: nonparametric estimation of the joint distribution function, modeling dependence with copulas and k-clustering for the exploratory analysis. Chapter 1 presents the general framework and the contributions of this thesis. Chapter 2 deals with the estimation of the joint distribution function of two censored variables in a simplified survival model in which the difference between two censoring variables is observed. We provide a new nonparametric estimator of the joint distribution function and we establish the asymptotic normality of the integrals with respect to its associated measure. Chapter 3 is devoted to nonparametric copula estimation under bivariate censoring. We provide a discrete and two smooth copula estimators along with two estimators of its density. The discrete estimator can be seen as an extension of the empirical copula under censoring. Chapter 4 provides a new exploratory approach for censored data analysis. We consider a multivariate configuration with one variable subjected to censoring and the others completely observed. We extend the probabilistic k-quantization method in the case of random vector with one censored component. The definitions of the empirical distortion and of empirically optimal quantizer are generalized in presence of one-dimensional censoring. We study the asymptotic properties of the distortion of the empirically optimal quantizer and we provide a non-asymptotic exponential bound for the rate of convergence. Our results are then applied to construct a new two-step clustering algorithm for censored data.