Dans le Chapitre 1, nous explorons le comportement joint des variables d'une marche aléatoire (X1, . . . ,Xn) lorsque leur valeur moyenne tend vers l'infini quand n tend vers l'infini. Il est prouvé que toutes ces variables doivent partager la même valeur, ce qui généralise les résultats précédents, dans le cadre de grands dépassements de sommes finies de i.i.d variables aléatoires. Dans le Chapitre 2, nous montrons un théorème de Gibbs conditionnel pour une marche aléatoire (X1, ..,Xn) conditionnée à une déviation extrême. Il est prouvé que lorsque les opérandes ont des queues légères avec une certaine régularité supplémentaire, la distribution asymptotique conditionnelle de X1 peut être approximée par la distribution tiltée en norme de la variation totale, généralisant ainsi le cas classique du LDP. Le troisième Chapitre explore le principe du maximum de vraisemblance dans les modèles paramétriques, dans le contexte du théorème de grandes déviations de Sanov. Le MLE est associé à la minimisation d'un critère spécifique de type divergence, qui se généralise au cas du bootstrap pondéré, où la divergnce est fonction de la distribution des poids. Certaines propriétés de la procédure résultante d'inférence sont présenteés ; l'efficacité de Bahadur de tests est également examinée dans ce contexte.

Competition between species is believed to lead to patterns of either competitive exclusion or limiting similarity within ecological communities; however, to date the amount of support for either as an outcome has been relatively weak. The two classes of null model commonly used to assess co-occurrence and limiting similarity have both been well studied for statistical performance; however, the methods used to evaluate their performance, particularly in terms of type II statistical errors, may have resulted in the underreporting of both patterns in the communities tested. The overall purpose of this study was to evaluate the efficacy of the negative co-occurrence and limiting similarity null models to detect patterns believed to result from competition between species and to develop an improved method for detecting said patterns. The null models were tested using synthetic but biologically realistic presence-absence matrices for both type I and type II error rate estimations. The effectiveness of the null models was evaluated with respect to community dimension (number of species × number of plots), and amount of pattern within the community. A novel method of subsetting species was developed to assess communities for patterns of co-occurrence and limiting similarity and four methods were assessed for their ability to isolate the species contributing signal to the pattern. Both classes of null model provided acceptable type I and type II error rates when matrices of more than 5 species and more than 5 plots were tested. When patterns of negative co-occurrence or limiting similarity were add to all species both null models were able to detect significant pattern (β > 0.95); however, when pattern was added to only a proportion of species the ability of the null models to detect pattern deteriorated rapidly with proportions of 80% or less. The use of species subsetting was able to detect significant pattern of both co-occurrence and limiting similarity when fewer than 80% of species were contributing signal but was dependent on the metric used for the limiting similarity null model. The ability of frequent pattern mining to isolate the species contributing signal shows promise; however, a more thorough evaluation is required in order to confirm or deny its utility.

Ce mémoire a pour objectif de déterminer si les précipitations convectives estivales simulées par le modèle régional canadien du climat (MRCC) sont stationnaires ou non à travers le temps. Pour répondre à cette question, nous proposons une méthodologie statistique de type fréquentiste et une de type bayésien. Pour l'approche fréquentiste, nous avons utilisé le contrôle de qualité standard ainsi que le CUSUM afin de déterminer si la moyenne a augmenté à travers les années. Pour l'approche bayésienne, nous avons comparé la distribution a posteriori des précipitations dans le temps. Pour ce faire, nous avons modélisé la densité \emph{a posteriori} d'une période donnée et nous l'avons comparée à la densité a posteriori d'une autre période plus éloignée dans le temps. Pour faire la comparaison, nous avons utilisé une statistique basée sur la distance d'Hellinger, la J-divergence ainsi que la norme L2. Au cours de ce mémoire, nous avons utilisé l'ARL (longueur moyenne de la séquence) pour calibrer et pour comparer chacun de nos outils. Une grande partie de ce mémoire sera donc dédiée à l'étude de l'ARL. Une fois nos outils bien calibrés, nous avons utilisé les simulations pour les comparer. Finalement, nous avons analysé les données du MRCC pour déterminer si elles sont stationnaires ou non. / The main goal of this master's thesis is to find whether the summer convective precipitations simulated by the Canadian Regional Climate Model (CRCM) are stationary over time or not. In order to answer that question, we propose both a frequentist and Bayesian statistical methodology. For the frequentist approach, we used standard quality control and the CUSUM to determine if the mean has increased over the years. For the Bayesian approach, we compared the posterior distributions of the precipitations over time. In order to do the comparison, we used a statistic based on the Hellinger's distance, the J-divergence and the L2 norm. In this master's thesis, we used the ARL (average run length) to calibrate each of our methods. Therefore, a big part of this thesis is about studying the actual property of the ARL. Once our tools are well calibrated, we used the simulation to compare them together. Finally, we studied the data from the CRCM to decide, whether or not, the data are stationary.

Sejam M uma variedade de Hadamard com curvatura seccional KM ≤ −k2 < 0 e ∂ M sua fronteira assintótica. Dizemos que M satisfaz a condição de convexidade estrita se, dados x ∈ ∂∞M e W ⊂ ∂∞M aberto relativo contendo x, existe um aberto Ω ⊂ M de classe C2 tais que x ∈ Int (∂ Ω) ⊂ W e M \ Ω ´e convexo. Provamos que a condição de convexidade estrita implica que M éregular no infinito com relação ao operador Q[u] := div a(|∇u|) \ |∇u| ∇u definido no espa¸co de Sobolev W 1,p(M ), onde a ∈ C1([0, +∞)) satisfaz a(0) = 0, at(s) > 0 para todo s > 0, a(s) ≤ C (sp−1 + 1), ∀s ≥ 0, onde C > 0 é uma constante, e a(s) ≥ sq para algum q > 0 e para s ≈ 0 e supomos que é possível resolver problemas de Dirichlet em bolas (compactas) de M com dados contínuos no bordo. Segue disto que sob a condição de convexidade estrita, os problemas de Dirichlet para equação de hipersuperfície mínima e para o p-laplaciano, p > 1, são solúveis para qualquer dado contínuo prescrito no bordo assintótico. Também provamos que se M é rotacionalmente simétrica ou se inf BR+1 KM ≥ −e 2kR /R2+2 , R ≥ R∗, para certos R∗ e E > 0, então M satisfaz a condição de convexidade estrita. / Let M be Hadamard manifold with sectional curvature KM ≤ −k2, k > 0 and ∂∞M its asymptotic boundary. We say that M satisfies the strict convexity condition if, given x ∈ ∂∞M and a relatively open subset W ⊂ 2 ∂∞M containing x, there exists a C open subset Ω ⊂ M such that x ∈ Int (∂∞Ω) ⊂ W and M \ Ω is convex. We prove that the strict convexity condition implies that M is regular at infinity relative to the operator Q [u] := div a(|∇u|) \ |∇u| ∇u , defined on the Sobolev space W 1,p(M ), where a ∈ C 1 ([0, ∞)) satisfies a(0) = 0, at(s) > 0 for all s > 0, a(s) ≤ C (s p−1 + 1), ∀s ≥ 0, where C > 0 is a constant, and a(s) ≥ sq , for some q > 0 and for s ≈ 0 and we suppose that it is possible to solve Dirichlet problems on (compact) balls of M with continuous boundary data. It follows that under the strict convexity condition, the Dirichlet problems for the minimal hypersurface and the p-Laplacian, p > 1, equations are solvable for any prescribed continuous asymptotic boundary data. We also prove that if M is rotationally symmetric or if inf BR+1 KM ≥ −e2kR/R2+2 , R ≥ R∗, for some R∗ and E > 0, then M satisfies the SC condition.

This study had as objective to estimate the genetic divergence among sugarcane genotypes by means of morphoagronomical characters and molecular markers, and to verify the relation between these procedures. An experiment was conducted in Rio Largo, AL, using a randomized block design with four repetitions. The multivariated analysis of Principal Components, the genetic divergence based on the Mahalanobis 2 ii' D Generalized Distance, and the Average Euclidean Distance Standardized were used for the analysis of the quantitative characters. Based on these distances, a grouping analysis was performed by the More Distant Neighbor method and the UPGMA method, besides Tocher for 2 ii' D . Jaccard coefficient and UPGMA grouping were used in the evaluation of the genetic divergence by molecular markers and morphologic characters. The inconsistency as to formation of different groups between the Standardized Average Euclidean Distance and 2 ii' D of Mahalanobis characterize these two estimates as measures of different dissimilarity. In the same way, the grouping techniques by the More Distant Neighbor method and by UPGMA show graphical dispersions that are not coincident, with differences in relation to the number of groups and in the grouping pattern, while the grouping by UPGMA and obtained by Tocher showed the same agreements. The 2 ii' D Distance of Mahalanobis corresponded to the Principal Components technique, by showing the same groups made by Tocher and UPGMA, obtained from 2 ii' D . However, these techniques did not indicate agreement with the Standardized Average Euclidean Distance. The correlation between the genetic divergence through morphologic characters and estimated by molecular markers was significant, however, from average magnitude (r = 0,47), indicating to be complementary measurements. There was not significant correlation for the divergence obtained through quantitative characteristics with the morphologic characters and molecular markers obtained by the different estimators, as well as between Standardized Average Euclidean Distance and the Mahalanobis 2 ii' D , which indicates clearly that there is no relation between these estimates. / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Este trabalho teve como objetivos a estimação da divergência genética entre genótipos de cana-de-açúcar, por meio de caracteres morfoagronômicos e marcadores moleculares, e verificar a relação entre esses procedimentos. Foi conduzido um experimento em Rio Largo, AL, utilizando o delineamento em blocos casualisados com quatro repetições. Para a análise dos caracteres quantitativos foi utilizada a análise multivariada de Componentes Principais, divergência genética a partir da Distância Generalizada 2 ii' D de Mahalanobis e Distância Euclidiana Média Padronizada. Com base nestas distâncias, realizaramse análise de agrupamento pelo método Hierárquico do Vizinho mais Distante e método UPGMA, além de Tocher para 2 ii' D . Na avaliação da divergência genética por meio de marcadores moleculares e caracteres morfológicos, utilizou-se o coeficiente de Jaccard e agrupamento UPGMA. As inconsistências quanto à formação de diferentes grupos entre a Distância Euclidiana Média Padronizada e a Distância 2 ii' D de Mahalanobis caracterizam estas duas estimativas como medidas de dissimilaridade distintas. Da mesma forma as técnicas de agrupamento pelo método do Vizinho mais Distante e por UPGMA evidenciam dispersões gráficas não coincidentes, com diferenças quanto ao número de grupos e ao padrão de agrupamento, ao passo que o agrupamento por UPGMA e obtido por Tocher apresentaram a mesma concordância. A Distância 2 ii' D de Mahalanobis correspondeu à técnica de Componentes Principais, por apresentarem os mesmos grupos formados por Tocher e UPGMA, obtidos a partir de 2 ii' D . No entanto, estas técnicas não indicaram concordância com a Distância Euclidiana Média Padronizada. A correlação entre a divergência genética através de caracteres morfológicos e a estimada por marcadores moleculares foi significativa, porém, de média magnitude (r = 0,47), indicando serem medidas complementares. Não houve correlação significativa para a divergência obtida por meio de características quantitativas com os caracteres morfológicos e marcadores moleculares, obtida pelos diferentes estimadores, assim como, entre a Distância Euclidiana Média Padronizada e a 2 ii' D de Mahalanobis, o que indica claramente não existir qualquer relação entre estas estimativas.

How to evaluate if a spatial model is well ajusted to a problem? How to know if it is the best model between the class of conditional autoregressive (CAR) and simultaneous autoregressive (SAR) models, including homoscedasticity and heteroscedasticity cases? To answer these questions inside Bayesian framework, we propose new ways to apply Bregman divergence, as well as recent information criteria as widely applicable information criterion (WAIC) and leave-one-out cross-validation (LOO). The functional Bregman divergence is a generalized form of the well known Kullback-Leiber (KL) divergence. There is many special cases of it which might be used to identify influential points. All the posterior distributions displayed in this text were estimate by Hamiltonian Monte Carlo (HMC), a optimized version of Metropolis-Hasting algorithm. All ideas showed here were evaluate by both: simulation and real data. / Como avaliar se um modelo espacial está bem ajustado? Como escolher o melhor modelo entre muitos da classe autorregressivo condicional (CAR) e autorregressivo simultâneo (SAR), homoscedásticos e heteroscedásticos? Para responder essas perguntas dentro do paradigma bayesiano, propomos novas formas de aplicar a divergência de Bregman, assim como critérios de informação bastante recentes na literatura, são eles o widely applicable information criterion (WAIC) e validação cruzada leave-one-out (LOO). O funcional de Bregman é uma generalização da famosa divergência de Kullback-Leiber (KL). Há diversos casos particulares dela que podem ser usados para identificar pontos influentes. Todas as distribuições a posteriori apresentadas nesta dissertação foram estimadas usando Monte Carlo Hamiltoniano (HMC), uma versão otimizada do algoritmo Metropolis-Hastings. Todas as ideias apresentadas neste texto foram submetidas a simulações e aplicadas em dados reais.

No description available.

This dissertation is an exploratory study aimed at empirically identifying and examining manifestations of communication barriers in the South African National Defence Force. A convergence model of communication is presented, describing the stages involved when individuals share the same information. The convergence model of communication delineates the relationship among the physical, psychological, and social aspects of communication. Two complementary theorems deduced by Kincaid (1988) from the convergence principle and the basic cybernetic process involved in information processing are applied in the analyses. The theorems specify the conditions under which intercultural communication will lead to convergence between members of different cultures involved. An empirically testable model was developed to collect information about possible communication barriers. The barriers are an indication of the restrictions on communication in the system and therefore indicate the possibility of restricting cultural convergence. / Communication / M.A. (Communication)

A identificação de atributos funcionais de colêmbolos associada com sua caracterização taxonômica responde aos impactos causados por mudanças de uso e manejo, além de terem um efeito sobre os processos funcionais do solo. Além disso, a descrição de atributos funcionais visa também atender o crescente interesse de investigar mecanismos relacionados com os processos de montagem de comunidade, tais como a convergência e divergência de atributos funcionais. Este trabalho teve três objetivos: 1) investigar a composição e estrutura de comunidades de colêmbolos em relação a alterações de uso do solo e diferentes ofertas de forragem através de uma abordagem funcional e taxonômica; 2) avaliar a existência de filtros ambientais e diferenciação de nichos dentro das comunidades; 3) avaliar o efeito de diferentes grupos funcionais dos colêmbolos no funcionamento do solo. Para isto, foram realizamos três estudos, sendo avaliados no primeiro estudo diferentes usos do solo (mata, campo e eucalipto) e no segundo estudo quatro níveis de ofertas de forragem. Como resultado principal de ambos os estudos, observamos que os de índices taxonômicos, tais como riqueza e diversidade de espécies, podem apresentar padrões diferentes dos índices funcionais, porém ambas as abordagens foram sensíveis aos tratamentos avaliados. De modo geral, o aumento da riqueza e diversidade de colêmbolos não é diretamente relacionado com o aumento da diversidade funcional. Também foi observado em ambos os estudos que a composição e estrutura das comunidades de colêmbolos foram afetadas significativamente pela existência de um filtro ambiental e pela diferenciação de nicho ecológico, uma vez que observamos convergência e divergência de atributos funcionais significativamente relacionados com os gradientes estudados. O terceiro estudo, realizado em condições de microcosmos, consistiram em 6 tratamentos e um controle, sendo os tratamento a combinação de diferentes grupos funcionais de colêmbolos. De modo geral, os resultados mostram que a presença de colêmbolos não afetou significativamente o tamanho da comunidade microbiana do solo, porém estimulou a mineralização de C e N. A presença de apenas um grupo funcional de colêmbolos estimulou a atividade microbiana, sendo estes significativamente diferentes do tratamento controle. Estas diferenças são ainda maiores quando há duas e três grupos funcionais no mesmo microcosmo. O estímulo sobre a comunidade microbiana possivelmente deve-se aos diferentes nichos ocupados pelos colêmbolos dentro do solo, pelas modificações na composição microbiana e pela redistribuição de nutrientes no interior do solo. Palavra-chave: colêmbolos; filtro ambiental; interação biológica; regras de montagem; campos; uso do solo; grupo funcional; atributo funcional; convergência de atributos; divergência de atributos. / Traits-based approach (functional traits) associated with taxonomical description in collembolan communities can response to environmental changes and have a directly effect on soil functioning. Furthermore, trait-based approach also aims to attend the growing interest in investigating mechanisms that are related to assembly process in communities, such as trait-convergence and trait-divergence assembly patterns. This work was three objective: 1) investigating the composition and structure of Collembola communities regarding environmental changes in soil and litter induced by land-use and grassland management using both taxonomic and functional approaches; 2) assessing the existence of environmental filtering and biotic interaction within collembolan communities; 3) evaluating the effect of three collembolan life-form on soil functioning. To this, we carried out three studies to investigate these objectives which two of them was carried out in three land-uses types (study 1) and four herbage allowance (study 2).The results of the studies 1 and 2 showed that taxonomic indices such as richness and taxonomical diversity and functional indices can have different patterns in collembolan communities, but both approaches are sensitive to the treatments evaluated. In general, increasing the richness and diversity of collembola was not directly related with the increase of the functional diversity. We also observed in both studies the composition and structure of Collembolan communities are significantly affected by the existence of an environmental filtering and biotic interaction, since we observed significantly trait-convergence and trait- divergence related to ecological gradients tested. Our findings show that functional traits provide a better understanding of the response of communities to biotic and abiotic changes. The third study was carried out in microcosm conditions and consisted of 6 different combinations of collembolan life-form. Overall, our results showed that collembolan life-forms did not significantly affect microbial biomass but collembolan life-forms affected microbial activities. The presence of only one life-form stimulated C and N mineralization regarding to their CO2 and enzymatic activities, but these stimulations were higher when there was two or three life-forms in the same microcosm. We suggest that stimulation on the microbial community is due to different niches occupied by each collembolan life-form, the changes on microbial composition and redistribution of nutrients in the soil.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This study aimed to verify whether there is genetic variability regarding nutritional and productivity characters of grains among cultivars of early cycle, very early cycle and transgenic maize and whether variability exists, discuss the differences among the cultivars, in order to select cultivars for crossing. It was used data of three experiments conducted during the harvest 2009/2010, at the experimental area of the Plant Science Department of the Federal University of Santa Maria. During these experiments were analyzed 76 maize cultivars, being 36 of early cycle, 22 of very early cycle and 18 of transgenic cultivar. In each experiment, after the harvest, in each of the three replicates of each cultivar, it was measured the following variables: grain productivity, crude protein, lysine, methionine, cysteine, threonine, tryptophan, valine, isoleucine, leucine, phenylalanine, histidine, arginine, ethere extract, starch and amylose in percentage of crude material. For each experiment, it was made an analysis of variance of each variable, according to a randomized block design. The averages of cultivars were compared by using Scott-Knott test. For each experiment was used the cluster analysis. Initially, discarding of variables was done in three steps: 1) Removal of variables without significant importance for the cultivar (F test of ANOVA), 2) Removal of variables that increase multicollinearity in the correlation matrix among the variablesand 3) Removal of variables by main components analysis. With the remaining variables was determined the Mahalanobis distance matrix (D2) and the cultivar clustering was done by using UPGMA method. Posteriorly, it was built a dendogram and calculated the cophenetic correlation coefficient. In order to test the hypothesis of differences among the average profile, for each experiment, it was calculated the evaluation of multivariate analysis of variance. There is variability among the early cycle, very early cycle and transgenic cultivars regarding grain productivity and nutritional characters in maize. Except starch which showed no variability in early cycle cultivar and transgenic cultivar, and methionine variables for transgenic cultivars. Based on variables grain productivity, crude protein and amylose, three groups were formed for early maturity cultivars, three groups for veryearly cycle cultivars and two groups for transgenic cultivars. / O presente trabalho teve como objetivos, verificar se há variabilidade genética, em relação aos caracteres nutricionais e à produtividade de grãos, entre as cultivares de ciclo precoce, de ciclo superprecoce e transgênicas de milho, e, existindo variabilidade, avaliar a divergência genética entre as cultivares com a finalidade de selecionar cultivares para cruzamentos. Foram utilizados os dados de três experimentos conduzidos na safra agrícola 2009/2010, na área experimental do Departamento de Fitotecnia, da Universidade Federal de Santa Maria. Nesses experimentos, foram avaliadas 76 cultivares de milho, sendo 36 de ciclo precoce, 22 de ciclo superprecoce e 18 cultivares transgênicas. Em cada experimento, após a colheita, em cada uma das três repetições de cada cultivar, foram mensuradas as seguintes variáveis: produtividade de grãos, proteína bruta, lisina, metionina, cisteina, treonina, triptofano, valina, isoleucina, leucina, fenilalanina, histidina, arginina, extrato etéreo, amido e amilose em porcentagem da matéria bruta (%MB). Para cada experimento, foi realizada a análise de variância de cada variável, conforme o modelo matemático de blocos ao acaso. As médias das cultivares foram comparadas por meio do teste de Scott-Knott. Para cada experimento foi realizada a análise de agrupamento. Para isso, inicialmente foi feito o descarte de variáveis em três etapas: 1) retirada de variáveis sem efeito significativo para cultivar (teste F da ANOVA); 2) retirada de variáveis causadoras de multicolinearidade na matriz de correlação entre as variáveis e; 3) descarte de variáveis por meio da análise de componentes principais. Após, com as variáveis que permaneceram. Com as variáveis que permaneceram foi determinada a matriz de distância de Mahalanobis (D2) e o agrupamento das cultivares foi realizado por meio do método UPGMA. Posteriormente, foi construído um dendrograma e calculado o coeficiente de correlação cofenética (CCC). Para testar a hipótese da diferença entre os perfis de médias de cada grupo, para cada experimento, foi realizada a análise de variância multivariada. Há variabilidade entre as cultivares de ciclo precoce, de ciclo superprecoce e cultivares transgênicas em relação à produtividade de grãos e aos caracteres nutricionais de grãos de milho. O amido não apresentou variabilidade em cultivares de ciclo precoce e cultivares transgênicas e a variável metionina para cultivares transgênicas. Com base nas variáveis produtividade de grãos, proteína bruta e amilose, foram formados três grupos para as cultivares de ciclo precoce, três grupos para as cultivares de ciclo superprecoce e dois grupos para as cultivares transgênicas.