Ce travail se situe dans le cadre de la régression non paramétrique univariée. Supposant la fonction de régression à variation bornée et partant du résultat selon lequel une telle fonction se décompose en la somme d’une fonction croissante et d’une fonction décroissante, nous proposons de construire et d’étudier un nouvel estimateur combinant les techniques d’estimation des modèles additifs et celles d’estimation sous contraintes de monotonie. Plus précisément, notreméthode consiste à itérer la régression isotonique selon l’algorithme backfitting. On dispose ainsià chaque itération d’un estimateur de la fonction de régression résultant de la somme d’une partiecroissante et d’une partie décroissante.Le premier chapitre propose un tour d’horizon des références relatives aux outils cités à l’instant. Le chapitre suivant est dédié à l’étude théorique de la régression isotonique itérée. Dans un premier temps, on montre que, la taille d’échantillon étant fixée, augmenter le nombre d’itérations conduit à l’interpolation des données. On réussit à identifier les limites des termes individuels de la somme en montrant l’égalité de notre algorithme avec celui consistant à itérer la régressionisotonique selon un algorithme de type réduction itérée du biais. Nous établissons enfin la consistance de l’estimateur.Le troisième chapitre est consacré à l’étude pratique de l’estimateur. Comme augmenter le nombre d’itérations conduit au sur-ajustement, il n’est pas souhaitable d’itérer la méthode jusqu’à la convergence. Nous examinons des règles d’arrêt basées sur des adaptations de critères usuellement employés dans le cadre des méthodes linéaires de lissage (AIC, BIC,...) ainsi que des critères supposant une connaissance a priori sur le nombre de modes de la fonction de régression. Il en ressort un comportement intéressant de la méthode lorsque la fonction de régression possède des points de rupture. Nous appliquons ensuite l’algorithme à des données réelles de type puces CGH où la détection de ruptures est d’un intérêt crucial. Enfin, une application à l’estimation des fonctions unimodales et à la détection de mode(s) est proposée / This thesis is part of non parametric univariate regression. Assume that the regression function is of bounded variation then the Jordan’s decomposition ensures that it can be written as the sum of an increasing function and a decreasing function. We propose and analyse a novel estimator which combines the isotonic regression related to the estimation of monotonefunctions and the backfitting algorithm devoted to the estimation of additive models. The first chapter provides an overview of the references related to isotonic regression and additive models. The next chapter is devoted to the theoretical study of iterative isotonic regression. As a first step we show that increasing the number of iterations tends to reproduce the data. Moreover, we manage to identify the individual limits by making a connexion with the general property of isotonicity of projection onto convex cones and deriving another equivalent algorithm based on iterative bias reduction. Finally, we establish the consistency of the estimator.The third chapter is devoted to the practical study of the estimator. As increasing the number of iterations leads to overfitting, it is not desirable to iterate the procedure until convergence. We examine stopping criteria based on adaptations of criteria usually used in the context of linear smoothing methods (AIC, BIC, ...) as well as criteria assuming the knowledge of thenumber of modes of the regression function. As it is observed an interesting behavior of the method when the regression function has breakpoints, we apply the algorithm to CGH-array data where breakopoints detections are of crucial interest. Finally, an application to the estimation of unimodal functions is proposed

Let t:[a,b] → ft be a C2 family of \"good\" C4 e piecewise expanding unimodal maps, with a critical point c, that is transversal to the topological classes of such maps. Given a lipchitzian observable ∅, consider the function ℛ∅(t)=∫∅dµt, where µt is the unique bsolutely continuous invariant probability of ft. We show a central limit theorem for the modulus of continuity of ℝ∅, that is limh→0m{t ∈ [a,b] : t + h ∈ [a,b] e 1/(Ψ(t)(-log|h|)½)((ℛ∅(t + h) - ℛ∅(t))/h) ≤ y} converges to 1/(2π)½ ∫y-∞e-s2/2ds. Now, let us consider a C2+ε expanding map f : 𝕊1 → 𝕊1 and a C1+ε periodic function v : 𝕊1 → ℝ. We show that the unique bounded solution of the twisted cohomological equation v(x) = α(f(x)) - Df(x)α(x) is either of class C1+ε or nowhere differentiable. We also prove that if α is nowhere differentiable, them the modulus of continuity of α satisfies a central limit theorem, that is, there is α &gt 0 such that limh→0µ{x : (α(x + h) - α(x))/(σ𝓁h(-log|h|)½) ≤ y} = 1/(2π)½ ∫y-∞e-t2/2dt, where µ is the absolutely continuous invariant probability of f. / Seja t:[a,b] → ft uma família C2 \"boa\" de transformações unimodais expansoras por pedaços com um ponto crítico c, que é transversal às classes topológicas de tais transformações. Dado um observável lipschitziano ∅, considere a função ℛ∅(t)=∫∅dµt, onde µt é a única probabiidade invariante absolutamente contínua de ft. Mostramos um teorema do limite central para o módulo de continuidade de ℝ∅, isto é limh→0m{t ∈ [a,b] : t + h ∈ [a,b] e 1/(Ψ(t)(-log|h|)½)((ℛ∅(t + h) - ℛ∅(t))/h) ≤ y} converge para 1/(2π)½ ∫y-∞e-s2/2ds. Vamos considerar agora f : 𝕊1 → 𝕊1 uma transformação expansora de classe C2+ε e v : 𝕊1 → ℝ uma função periódica de classe C1+ε. Mostramos que a única solução limitada da equação cohomológica torcida v(x) = α(f(x)) - Df(x)α(x) ou é de classe C1+ε ou não possui derivada em ponto algum. Mostramos também que se α não possui derivada em ponto algum, então o módulo de continuidade de α satisfaz um teorema do limite central, isto é, existe α &gt 0 tal que limh→0µ{x : (α(x + h) - α(x))/(σ𝓁h(-log|h|)½) ≤ y} = 1/(2π)½ ∫y-∞e-t2/2dt, onde µ é a probabilidade invariante absolutamente contínua associada a f.

Background: It is widely believed that the median of a unimodal distribution is "usually" between the mean and the mode for right skewed or left skewed distributions. However, this is not always true, especially with grouped data. For some research, analyses must be conducted based on grouped data since complete raw data are not always available. A gap exists in the body of research on the mean-median-mode inequality for grouped data. Methods: For grouped data, the median Me=L+((n/2-F)/fm)×d and the mode Mo=L+(D1/(D1+D2))×d, where L is the median/modal group lower boundary, n is the total frequency, F and G are the cumulative frequencies of the groups before and after the median/modal group respectively, D1= fm - fm-1 and D2=fm - fm+1, fmis the median/modal group frequency, fm-1 and fm+1 are the premodal and postmodal group frequency respectively. Assuming there are k groups and k is odd, group width d is the same for each group and the mode and median are within (k+1)/2th group. Necessary and sufficient conditions are derived for each of six arrangements of mean, median and mode. Results: Table available at https://apha.confex.com/apha/143am/webprogram/Paper326538.html Conclusion: For grouped data, the mean-median-mode inequality can be any order of six possibilities.

Graphs provide interesting ways to generate families of lattice polytopes. In particular, one can use matrices encoding the information of a finite graph to define vertices of a polytope. This dissertation initiates the study of the Laplacian simplex, PG, obtained from a finite graph G by taking the convex hull of the columns of the Laplacian matrix for G. The Laplacian simplex is extended through the use of a parallel construction with a finite digraph D to obtain the Laplacian polytope, PD. Basic properties of both families of simplices, PG and PD, are established using techniques from Ehrhart theory. Motivated by a well-known conjecture in the field, our investigation focuses on reflexivity, the integer decomposition property, and unimodality of Ehrhart h*-vectors of these polytopes. A systematic investigation of PG for trees, cycles, and complete graphs is provided, which is enhanced by an investigation of PD for cyclic digraphs. We form intriguing connections with other families of simplices and produce G and D such that the h*-vectors of PG and PD exhibit extremal behavior.

The semantic content (i.e., meaning of words) is the essence of retrieved autobiographical memories. In comparison to previous research, which has mainly focused on phenomenological experiences and age distribution of memory events, the present study provides a novel view on the retrieval of event information by addressing the semantic representation of memories. In the present study the semantic representation (i.e., word locations represented by vectors in a high dimensional space) of retrieved memory information were investigated, by analyzing the data with an automatic statistical algorithm. The experiment comprised a cued recall task, where participants were presented with unimodal (i.e., one sense modality) or multimodal (i.e., three sense modalities in conjunction) retrieval cues and asked to recall autobiographical memories. The memories were verbally narrated, recorded and transcribed to text. The semantic content of the memory narrations was analyzed with a semantic representation generated by latent semantic analysis (LSA). The results indicated that the semantic representation of visually evoked memories were most similar to the multimodally evoked memories, followed by auditorily and olfactorily evoked memories. By categorizing the semantic content into clusters, the present study also identified unique characteristics in the memory content across modalities.

The maximum spacing (MSP) method, introduced by Cheng and Amin (1983) and independently by Ranneby (1984), is a general estimation method for continuous univariate distributions. The MSP method, which is closely related to the maximum likelihood (ML) method, can be derived from an approximation based on simple spacings of the Kullback-Leibler information. It is known to give consistent and asymptotically efficient estimates under general conditions and works also in situations where the ML method fails, e.g. for the three parameter Weibull model. In this thesis it is proved under general conditions that MSP estimates of parameters in the Euclidian metric are strongly consistent. The ideas behind the MSP method are extended and a class of estimation methods is introduced. These methods, called generalized MSP methods, are derived from approxima­tions based on sum-functions of rath order spacings of certain information mea­sures, i.e. the ^-divergences introduced by Csiszår (1963). It is shown under general conditions that generalized MSP methods give consistent estimates. In particular, it is proved that generalized MSP methods give L1 consistent esti­mates in any family of distributions with unimodal densities, without any further conditions on the distributions. Other properties such as distributional robust­ness are also discussed. Several limit theorems for sum-functions of rath order spacings are given, for ra fixed as well as for the case when ra is allowed to in­crease to infinity with the sample size. These results provide a strongly consistent nonparametric estimator of entropy, as well as a characterization of the uniform distribution. Further, it is shown that Cressie's (1976) goodness of fit test is strongly consistent against all continuous alternatives. / digitalisering@umu

Let t:[a,b] → ft be a C2 family of \"good\" C4 e piecewise expanding unimodal maps, with a critical point c, that is transversal to the topological classes of such maps. Given a lipchitzian observable ∅, consider the function ℛ∅(t)=∫∅dµt, where µt is the unique bsolutely continuous invariant probability of ft. We show a central limit theorem for the modulus of continuity of ℝ∅, that is limh→0m{t ∈ [a,b] : t + h ∈ [a,b] e 1/(Ψ(t)(-log|h|)½)((ℛ∅(t + h) - ℛ∅(t))/h) ≤ y} converges to 1/(2π)½ ∫y-∞e-s2/2ds. Now, let us consider a C2+ε expanding map f : 𝕊1 → 𝕊1 and a C1+ε periodic function v : 𝕊1 → ℝ. We show that the unique bounded solution of the twisted cohomological equation v(x) = α(f(x)) - Df(x)α(x) is either of class C1+ε or nowhere differentiable. We also prove that if α is nowhere differentiable, them the modulus of continuity of α satisfies a central limit theorem, that is, there is α &gt 0 such that limh→0µ{x : (α(x + h) - α(x))/(σ𝓁h(-log|h|)½) ≤ y} = 1/(2π)½ ∫y-∞e-t2/2dt, where µ is the absolutely continuous invariant probability of f. / Seja t:[a,b] → ft uma família C2 \"boa\" de transformações unimodais expansoras por pedaços com um ponto crítico c, que é transversal às classes topológicas de tais transformações. Dado um observável lipschitziano ∅, considere a função ℛ∅(t)=∫∅dµt, onde µt é a única probabiidade invariante absolutamente contínua de ft. Mostramos um teorema do limite central para o módulo de continuidade de ℝ∅, isto é limh→0m{t ∈ [a,b] : t + h ∈ [a,b] e 1/(Ψ(t)(-log|h|)½)((ℛ∅(t + h) - ℛ∅(t))/h) ≤ y} converge para 1/(2π)½ ∫y-∞e-s2/2ds. Vamos considerar agora f : 𝕊1 → 𝕊1 uma transformação expansora de classe C2+ε e v : 𝕊1 → ℝ uma função periódica de classe C1+ε. Mostramos que a única solução limitada da equação cohomológica torcida v(x) = α(f(x)) - Df(x)α(x) ou é de classe C1+ε ou não possui derivada em ponto algum. Mostramos também que se α não possui derivada em ponto algum, então o módulo de continuidade de α satisfaz um teorema do limite central, isto é, existe α &gt 0 tal que limh→0µ{x : (α(x + h) - α(x))/(σ𝓁h(-log|h|)½) ≤ y} = 1/(2π)½ ∫y-∞e-t2/2dt, onde µ é a probabilidade invariante absolutamente contínua associada a f.

This dissertation is a study of the dynamics of one-dimensional unimodal maps and is mainly concerned with those maps which are trapezoidal. The trapezoidal function, f_e, is defined for eΣ(0,1/2) by f_e(x)=x/e for xΣ[0,e], f_e(x)=1 for xΣ(e,1-e), and f_e(x)=(1-x)/e for xΣ[1-e,1]. We study the symbolic dynamics of the kneading sequences and relate them to the analytic dynamics of these maps. Chapter one is an overview of the present theory of Metropolis, Stein, and Stein (MSS). In Chapter two a formula is given that counts the number of MSS sequences of length n. Next, the number of distinct primitive colorings of n beads with two colors, as counted by Gilbert and Riordan, is shown to equal the number of MSS sequences of length n. An algorithm is given that produces a bisection between these two quantities for each n. Lastly, the number of negative orbits of size n for the function f(z)=z^2-2, as counted by P.J. Myrberg, is shown to equal the number of MSS sequences of length n. For an MSS sequence P, let H_ϖ(P) be the unique common extension of the harmonics of P. In Chapter three it is proved that there is exactly one J(P)Σ[0,1] such that the itinerary of λ(P) under the map is λ(P)f_e is H_ϖ(P). In Chapter four it is shown that only period doubling or period halving bifurcations can occur for the family λf_e, λΣ[0,1]. Results concerning how the size of a stable orbit changes as bifurcations of the family λf_e occur are given. Let λΣ[0,1] be such that 1/2 is a periodic point of λf_e. In this case 1/2 is superstable. Chapter five investigates the boundary of the basin of attraction of this stable orbit. An algorithm is given that yields a graph directed construction such that the object constructed is the basin boundary. From this we analyze the Hausdorff dimension and measure in that dimension of the boundary. The dimension is related to the simple β-numbers, as defined by Parry.

It is widely believed that the median is “usually” between the mean and the mode for skewed unimodal distributions. However, this inequality is not always true, especially with grouped data. Unavailability of complete raw data further necessitates the importance of evaluating this characteristic in grouped data. There is a gap in the current statistical literature on assessing mean–median–mode inequality for grouped data. The study aims to evaluate the relationship between the mean, median, and mode with unimodal grouped data; derive conditions for their inequalities; and present their application.

Effect of architecture at nanoscale on the macroscopic properties of polymer materials has long been a field of major interest, as evidenced by inhomogeneities in networks, multimodal network topologies, etc. The primary purpose of this research is to establish the architecture-property relationship of polymer networks by studying the physical and mechanical responses of a series of topologically different PTHF networks. Monodispersed allyl-terminated PTHF precursors were synthesized through ¡°living¡± cationic polymerization and functional end-capping. Model networks of various crosslink densities and inhomogeneities levels (unimodal, bimodal and clustered) were prepared by endlinking precursors via thiol-ene reaction. Thermal characteristics, i.e., glass transition, melting point, and heat of fusion, of model PTHF networks were investigated as functions of crosslink density and inhomogeneities, which showed different dependence on these two architectural parameters. Study of freezing point depression (FPD) of solvent confined in swollen networks indicated that the size of solvent microcrystals is comparable to the mesh size formed by intercrosslink chains depending on crosslink density and inhomogeneities. Relationship between crystal size and FPD provided a good reflection of the existing architecture facts in the networks. Mechanical responses of elastic chains to uniaxial strains were studied through SANS. Spatial inhomogeneities in bimodal and clustered networks gave rise to ¡°abnormal butterfly patterns¡±, which became more pronounced as elongation ratio increases. Radii of gyration of chains were analyzed at directions parallel and perpendicular to stretching axis. Dependence of Rg on ¦Ë was compared to three rubber elasticity models and the molecular deformation mechanisms for unimodal, bimodal and clustered networks were explored. The thesis focused its last part on the investigation of evolution of free volume distribution of linear polymer (PE) subjected to uniaxial strain at various temperatures using a combination of MD, hard sphere probe method and Voronoi tessellation. Combined effects of temperature and strain on free volume were studied and mechanism of formation of large and ellipsoidal free volume voids was explored.