Bayesian Model Averaging (BMA) is a new area in modern applied statistics that provides data analysts with an efficient tool for discovering promising models and obtaining esti-mates of their posterior probabilities via Markov chain Monte Carlo (MCMC). These probabilities can be further used as weights for model averaged predictions and estimates of the parameters of interest. As a result, variance components due to model selection are estimated and accounted for, contrary to the practice of conventional data analysis (such as, for example, stepwise model selection). In addition, variable activation probabilities can be obtained for each variable of interest. This dissertation is aimed at connecting BMA and various ramifications of the multivari-ate technique called Reduced-Rank Regression (RRR). In particular, we are concerned with Canonical Correspondence Analysis (CCA) in ecological applications where the data are represented by a site by species abundance matrix with site-specific covariates. Our goal is to incorporate the multivariate techniques, such as Redundancy Analysis and Ca-nonical Correspondence Analysis into the general machinery of BMA, taking into account such complicating phenomena as outliers and clustering of observations within a single data-analysis strategy. Traditional implementations of model averaging are concerned with selection of variables. We extend the methodology of BMA to selection of subgroups of observations and im-plement several approaches to cluster and outlier analysis in the context of the multivari-ate regression model. The proposed algorithm of cluster analysis can accommodate re-strictions on the resulting partition of observations when some of them form sub-clusters that have to be preserved when larger clusters are formed. / Ph. D.

Ένα από τα πιο αποτελεσματικά μέσα για την επιτυχία ενός προϊόντος είναι αναμφισβήτητα η Έρευνα Αγοράς (Market Research). Η έρευνα υλοποιείται σε αντιπροσωπευτικό δείγμα ενός συγκεκριμένου υπό μελέτη πληθυσμού, με απώτερο σκοπό να εξαχθούν συμπεράσματα από την ποσοτική στατιστική επεξεργασία και ανάλυση των δεδομένων σχετικά με τις απόψεις, τις στάσεις και τις συμπεριφορές που διαθέτουν οι καταναλωτές, τις τάσεις και τις προοπτικές ενός κλάδου, το μερίδιο αγοράς που αντιστοιχεί σε κάποιο προϊόν ή επίσης και τη γνώση που έχουν οι καταναλωτές όσον αφορά τη μάρκα ή την εικόνα ενός προϊόντος. Η στατιστική ανάλυση των αποτελεσμάτων επιτυγχάνεται με χρήση διαφόρων στατιστικών τεχνικών, ανάμεσα στις οποίες περιλαμβάνεται και η Ανάλυση Αντιστοιχιών (Correspondence Analysis). Η Ανάλυση Αντιστοιχιών (Correspondence Analysis) είναι μια διερευνητική τεχνική ανάλυσης δεδομένων που απεικονίζει γραφικά πίνακες συνάφειας και πολυμεταβλητά κατηγορικά δεδομένα, ενώ παράλληλα τις τελευταίες δεκαετίες έχει σημειώσει ταχύτατη εξέλιξη και ανάπτυξη. Η Ανάλυση Αντιστοιχιών βρίσκει εφαρμογή τόσο στην Έρευνα Αγοράς, ακριβώς λόγω της κατηγορικής φύσεως των δεδομένων που προκύπτουν σε έρευνες τέτοιου τύπου, όσο και γενικότερα στις κοινωνικές επιστήμες όπου συνήθως χρειάζεται ένα στατιστικό «εργαλείο» για να ελέγξει τις αντιστοιχίες- συσχετίσεις μεταξύ των μεταβλητών και να τις αποδώσει με τη βοήθεια ενός γραφήματος. Σκοπός της παρούσας εργασίας είναι η θεωρητική, πρακτική και υπολογιστική επισκόπηση της συγκεκριμένης στατιστικής μεθόδου καθώς και η υλοποίησή της στα στατιστικά πακέτα SPSS & MINITAB, με τη χρήση κατάλληλων δεδομένων από τον τομέα της Έρευνας Αγοράς. Θα δοθεί έμφαση περισσότερο στην ερμηνεία των αποτελεσμάτων παρά στις μαθηματικές και τεχνικές λεπτομέρειες των διαδικασιών της εφαρμογής. / One of the most effective means for the success of a product is indisputably the Market Research. The research is conducted in a representative sample of a specific population under study with the aim to draw conclusions which are derived from the quantitative statistical processing and analysis of data on the opinions, the attitudes and behaviors that consumers have, the trends and the prospects of a branch, the market share which is corresponding to a product or also the knowledge that consumers have got as far as it concerns the brand or product image. The statistical analysis of the results is accomplished by using different statistical techniques including Correspondence Analysis. The Correspondence Analysis is an exploratory data analysis technique that depicts graphically contingency tables and multivariate categorical data, while the last decades it has made rapid progress and development. The Correspondence Analysis is applicable to Market Research, just because of the categorical nature of data obtained in researches of this type and generally in social sciences, where commonly a statistical tool is needed to check on the correspondences- correlations among the variables and yield them through a graph. The purpose of this study is the theoretical, practical and computational survey of this specific statistical method and its implementation in the statistical packages SPSS & MINITAB by using appropriate data from Market Research. Emphasis will be given more on the interpretation of results despite the mathematical and technical details of the method procedures.

Η παρούσα διπλωματική εργασία ασχολείται με πολυδιάστατα κατηγορικά δεδομένα όπως αυτά προκύπτουν από συλλογή μέσω ερωτηματολογίων. Για να αναλυθεί όμως ένα ερωτηματολόγιο το οποίο περιλαμβάνει πλήθος ερωτήσεων-μεταβλητών και να εξάγουμε ορισμένα χρήσιμα συμπεράσματα θα πρέπει, πρώτα από όλα να γίνει η κατάλληλη κωδικοποίηση των δεδομένων. Χρησιμοποιώντας στατιστικές τεχνικές και μεθόδους κατάλληλες για κατηγορικά δεδομένα μπορούμε πιο εύκολα να μελετήσουμε τις σχέσεις μεταξύ των μεταβλητών. Για τον σκοπό αυτό, παρουσιάζουμε και αναλύουμε τη θεωρία της Παραγοντικής Ανάλυσης των Αντιστοιχιών και της Πολλαπλής Παραγοντικής Ανάλυσης των Αντιστοιχιών. Ύστερα, αναλύουμε τα αποτελέσματα που δίνουν οι μέθοδοι όταν τις χρησιμοποιήσουμε για την ανάλυση του ερωτηματολογίου. Τα συμπεράσματα που προκύπτουν έχουν ιδιαίτερο ενδιαφέρον. Η δυναμικότητα των μοντέλων αυτών παρουσιάζεται μέσα από μία εφαρμογή από τον χώρο των κοινωνικών επιστημών σε θέματα που αφορούν τα σχολεία και τους μαθητές της Πρωτοβάθμιας εκπαίδευσης. Στις μεθόδους αυτές δίνεται έμφαση κυρίως στα γραφικά αποτελέσματα αλλά και στις εκτιμήσεις των σκορ των κατηγοριών των μεταβλητών. Όλα τα παραπάνω τα συγκρίνουμε κριτικά μεταξύ τους στη θεωρία και στη πράξη έτσι ώστε ο ενδιαφερόμενος αναγνώστης να κατανοήσει περισσότερο τις μεθόδους αυτές και να αποκομίσει όσο το δυνατόν περισσότερες πληροφορίες που θα τον βοηθήσουν για την εφαρμογή τους. / This dissertation deals with multivariate categorical data of a raw data set produced by a questionnaire designed for a research purpose. However, in order to analyze a questionnaire and extract some fruitful results, that includes a great number of questions-variables, we must first impose a structure on it especially on situations this specific structure is missing. Whenever the structure is imposed, by using statistical techniques and methods designed for categorical data, we can then study more efficiently the relations among the variables in concern for further analyses. The capacity of these models is presented through an application from the social sciences on issues concerning schools and pupils in primary education. The analysis on a smaller subset is further explored by describing the issues of Correspondence Analysis, and Multiple Correspondence Analysis. With these methods, we focus on the interpretation of the results on the graphical displays of the data but also on the estimated category scores of the variables. The above methods described in this dissertation and the results after implementing them are all critically compared with each other at each chapter. This gives to the interesting reader the possibility to fully understand them and to obtain additional information on their implementation.

The aim of this thesis is to introduce statistical method called Correspondence analysis as a strong instrument for exploratory data analysis. The main purpose is to understand how to interpret the correspondence map, the graphical output of this method, correctly. The method is presented both in its simple version, and its extension to multivariate data. Usage of method is demonstrated on data of non-financial subjects of Czech republic, which are entered in the register of economic subjects.

The vegetation responses to seven silvicultural treatments one growing season after harvesting were examined on seven sites in the southern Appalachian mountains of Virginia and West Virginia. Treatments included: 1) control, 2) understory control by herbicide, 3) group selection, 4) high-leave shelterwood, 5) low-leave shelterwood, 6) leave tree, and 7) clearcut. The effects of harvesting were compared between treatments and between pre-harvest and post-harvest samplings. Species richness, percent cover, and local species extinctions were calculated for sample plots ranging in size from 1m2 to 2 ha. Vegetation richness and cover increased with increasing harvest intensity. Local species extinctions were similar in the control and disturbed treatments. Additional analyses were performed using the control, high-leave shelterwood, and clearcut on five of the seven sites to determine the relationships between soil, litter, and other environmental characteristics and vegetation in the herbaceous layer (<1 m in height). Multivariate analysis techniques were used to analyze average differences in species abundance between pre-harvest and post-harvest and to relate post-harvest vegetation to microsite characteristics. Regional-scale differences in site location were more important in explaining the presence of a species than were environmental characteristics. Within a region, species primarily were distributed along a light/litter weight gradient and secondarily along a soil properties and nutrient gradient. / Master of Science

In the mid-10th century the import of dirhems from Eastern Europe came to an end. From being a steady flow of Islamic coins from Viking-Age Russia the focus shifted to western European coins such as German and English. Most studies on Viking Age coins and hoards in general have focused on determine the individual coin type and the composition of the hoard in whole. Others have focused on the more social aspects on why the hoards even exist. My study will investigate if it is possible to detect if there are different networks behind the import of coins by analysing the composition of hoards in the transitional phase. By using a correspondence analysis and also GIS analysis, I want to investigate differences and similarities in the hoards composition. The correspondence analysis will show if there are any affiliations between different variables, which a means recurring pattern or combination of data represented in the hoards such as mints, size, TPQ and find distribution on Gotland etc. These presence or absence of affiliations will then be discussed if they could indicate different networks operating the coin import. The material the study is based on are 10th century hoards from Gotland which contain coins from late Islamic dynasties, Germany, England and the Byzantine empire. The result shows no specific indications in the hoards composition on different networks or groups that imported different kinds of coins. There is however a chance that different local groups have imported the same type of coins. The correspondence analysis also clearly visualizes the shift from the import of Islamic dirhams to Western coins.

Tidal freshwater forest restoration after dam removal has been unexplored to date. This study elucidated pre-dam forest composition, as well as post-dam edaphic and microtopographical attributes and woody species recruiting along a narrow ecotone of a 29.3-ha tidal freshwater wetland. The ≈65-year-old historical forest (15 species, 200 stems ha-1) and ≈7-year-old contemporary forest (40 species and 11,009 stems ha-) community dominants were dissimilar (Fraxinus spp. vs. Liquidambar styraciflua, respectively). Pre-dam environmental conditions were unknown. Post-dam edaphic water content, organic matter, redox potential and microtopography differed significantly across tidal sites but were less variable in non-tidal sites. Shifts in the contemporary woody community composition and the concomitant increase in stem density and seedling:sapling ratios with elevation likely owed to significant changes in microtopography and edaphic attributes. Developing ecotones that contain variable microtopography may be extremely important for successful natural woody recruitment after dam removal from a tidal freshwater system.

Orientador: Maria Rita Marques de Oliveira / Banca: Juliana Álvares Duarte Bonini Campos / Banca: Maria Márcia Pereira Sartori / Resumo: A avaliação do padrão alimentar vem emergindo como uma alternativa nos estudos de epidemiologia nutricional, substituindo as pesquisas baseadas em análise de nutrientes isolados. No entanto, a subnotificação da ingestão, que é fortemente evidenciada entre indivíduos com excesso de peso, influencia a avaliação do consumo energético e pode, consequentemente, induzir a um padrão de consumo alimentar diferente quando comparado ao padrão de consumo de pessoas notificadoras plausíveis. Uma vez que estudos sobre o padrão alimentar ainda não foram avaliados sob a ótica da subnotificação, este trabalho teve por objetivo avaliar a influência da subnotificação do consumo no padrão alimentar, discriminado por grupos alimentares e por nutrientes específicos, entre mulheres obesas da fila de espera da cirurgia bariátrica. Os resultados obtidos foram discutidos em dois artigos científicos, sendo que o primeiro evidenciou a diferença nos padrões alimentares discriminados por contribuição energética diária dos subgrupos alimentares entre as mulheres notificadoras e subnotificadoras do consumo alimentar, e o segundo evidenciou as diferenças entre os grupos de mulheres para os padrões de consumo e itens dos subgrupos alimentares, ambos discriminados pelos nutrientes. Participaram do estudo 412 candidatas à cirurgia bariátrica da Clínica Bariátrica de Piracicaba-SP, Brasil. Dados referentes às informações gerais, e medidas antropométricas foram coletados e utilizados para os cálculos preditivos de Índice de Massa Corporal (IMC), Gasto Energético de Repouso (GER) e Necessidade Energética Total (NET). As informações de consumo alimentar e Nível de Atividade Física (NAF) preditivos foram utilizadas na classificação das mulheres quanto a plausibilidade da notificação do consumo alimentar, confrontando... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Assessment of food patterns has emerged as an alternative in nutritional epidemiology studies, replacing studies based on analysis of individual nutrients. However, food intake underreporting, very common among individuals with excess weight, affects the assessment of energy intake and may, consequently, induce a food intake pattern different from that of plausible energy reporters. Since food pattern studies have not yet been done from the underreporting viewpoint, the objective of this study was to assess the influence of underreporting on the specific food groups and nutrients consumed by obese women waiting for bariatric surgery. The results were discussed in two scientific articles. The first article evidenced the difference between the food patterns of reporters and underreporters according to the daily energy contribution of different food subgroups. The second article evidenced the differences in intake patterns and items in food subgroups, both according to nutrients, between the two groups of women. A total of 412 bariatric surgery candidates of the Bariatric Clinic of Piracicaba, SP, Brazil, participated in the study. General information and anthropometric measurements were collected and used for calculating body mass index (BMI) and predicting resting energy expenditure (REE) and total energy requirement (TER). Food intake and physical activity level (PAL) were used for classifying women regarding the plausibility of the reported food intake, confronting the ratio of reported energy intake EIrep to REE with PAL (EIrep:REE=PAL), considering the variances of this equation's components. Food subgroups were determined by the Brazilian Food Guide. The food patterns of both groups of women were discriminated according to the food subgroups predefined by exploratory... (Complete abstract click electronic access below) / Mestre

Avec l'accroissement du nombre de génomes séquencés, l'organisation de ces données brutes et des données dérivées, l'extraction de l'information et des connaissances associées défie l'imagination. La notion de voisinage a été d'abord été introduite pour l'organisation des données dans des bases de données relationnelles. Pour extraire des informations pertinentes à partir de données massives, différents types de voisinages ont été étudiés ici. Tout d'abord, avec l'analysedes correspondances (CA) et en utilisant le regroupement supervisé ("model clustering" MBC), la proximité mutuelle des éléments formant deux entités biologiques centrales, les gènes (codant les protéines) et les acides aminés a été analysée. Nous montrons par exemple que les protéines de Psychromonas ingrahamii, bactérie psychrophile extrêmes, sont regroupées en six classes, et qu'il y a une forte opposition entre le comportement de l'asparagine (N) et des acides aminés sensibles à l'oxygène, ce que nous expliquons en terms de résistance au froid. Ensuite, nous avons analysé la répartition entre les îlots génomiques (GI) et le squelette du génome de base à partir d'une nouvelle méthode combinant composition en bases et en gènes, caractéristiques GI et de briser les synténies. L'application de cette approche à E. coli et B. subtilis a révélé que cette nouvelle méthode permet d'extraire certaines régions significative, non publiées auparavant.Enfin, pour illustrer un voisinage fin, la régulation de l'expression d'un gène et son évolution, nous avons étudié la relation entre les régions en amont du gène et la zone codante du gène thrS de façon approfondie. Nous avons constaté que ces deux régions associées à un gène, se sont comportés différemment dans l'histoire évolutive. Certaines des régions en amont porteuses de la fonction non-essentielle de régulation (qui contrôle l'expression de gène) ont évoluédifféremment de la région codante. / With more and more genomes being sequenced, the organization of those raw data and the derived data, the extraction of information and knowledge from these data has become a challenge. A key concept in this field is that of the neighborhood, especially with respect to the organization of data in relational databases. To extract information from bulk data, different kinds of neighborhoods were studied and each show interesting results in current study. .Firstly, through the Correspondence Analysis (CA) and later Model Based Clustering (MBC), two kinds of neighbors i.e. the genes (proteins) and amino acids were analyzed respectively, and it was found that proteins from Psychromonas ingrahamii are clustered into six classes, and there is strong opposition between asparagine (N) and the oxygen-sensitive amino acids. Secondly, the relationship between genomic islands and core genome (i.e. two closely linked neighbors withlarge range on the chromosome) was studied by a new method combining composition, GI features and synteny break. On applying to E. coli and B. subtilis it was revealed that this new method can extract some meaningful regions not published before. Thirdly, the relationship between upstream and coding regions of thrS gene (i.e. a case for two closely linked neighbors with small range on the chromosome) was studied extensively. It was found that these two regions associated to one gene, behaved differently in the evolutionary history.. Some of the upstream regions bearing non-essential function (i.e. regulation of gene expression) evolved more slowly than the coding region.

This thesis uses the results of limited testing at the Fort Dade (1836-1842) and Hope Homestead (1842-ca. 1900) archaeological sites to establish a method for quickly identifying threatened sites with minimal disturbance to surrounding areas. Artifact analysis, pattern recognition, and comparison with similar known sites are key elements. Pedestrian survey, metal detection, posthole and shovel testing, and test excavation are tools used to accomplish this. Artifact analysis is used to establish date ranges for the sites, as well as the material variation between military and homestead occupations. Artifacts used for analysis include glass, ceramics, nails, arms and personal items. Quantitative analysis of artifact assemblages is utilized to determine broad site type classification, and further contribute to preliminary identification. Correspondence analysis helps differentiate sites according to length and type of occupation. With refinement, this method could be used for preliminary identification of many Seminole War sites. Rapid and widespread development in Florida has made identification of Seminole War sites a priority, so they can be recorded and preserved before they are lost forever.