本論文所探討之主題乃是針對一般所利用之集群方法，試著以整數規劃的方法來探討 集群分析的問題。 整數規劃之特性在於其所得之分組結果為真正的最佳解，而一般集群方法（如連鎖法 ，k 一平均數法）所得之結果僅是局部最佳解。 本文共分五章，第一章為緒論；第二章簡介一般集群方法；第三章建立四個整數規劃 的模型，俾用以解決不同需求之集群分析的問題；第四章實例探討，以某國中學生之 學科成績做為集群分析之變數，將每個學生依其成績高低而予以分組，並就一般集群 方法及整數規劃方法各作分析，並予比較；第五章為結論。全文共計一冊，約一萬五 仟字。

The Bayesian approach to cluster analysis is presented. We assume that all data stem from a finite mixture model, where each component corresponds to one cluster and is given by a multivariate normal distribution with unknown mean and variance. The method produces posterior distributions of all cluster parameters and proportions as well as associated cluster probabilities for all objects. We extend this method in several directions to some common but non-standard situations. The first extension covers the case with a few deviant observations not belonging to one of the normal clusters. An extra component/cluster is created for them, which has a larger variance or a different distribution, e.g. is uniform over the whole range. The second extension is clustering of longitudinal data. All units are clustered at all time points separately and the movements between time points are modeled by Markov transition matrices. This means that the clustering at one time point will be affected by what happens at the neighbouring time points. The third extension handles datasets with missing data, e.g. item non-response. We impute the missing values iteratively in an extra step of the Gibbs sampler estimation algorithm. The Bayesian inference of mixture models has many advantages over the classical approach. However, it is not without computational difficulties. A software package, written in Matlab for Bayesian inference of mixture models is introduced. The programs of the package handle the basic cases of clustering data that are assumed to arise from mixture models of multivariate normal distributions, as well as the non-standard situations.

Cette thèse poursuit deux objectifs principaux. Le premier consiste à développer un modèle de la compatibilité conjugale fondé sur les notions d’attachement (évitement de la proximité et anxiété d’abandon) et de personnalité (névrotisme). Le deuxième objectif est d’examiner la validité prévisionnelle de ce modèle de compatibilité chez des couples en psychothérapie conjugale. Dans un premier temps, la présente thèse fait brièvement état de la situation des couples au Québec et des répercussions de la détresse conjugale. Par la suite, la recension de la documentation traite du concept de compatibilité entre conjoints en tant qu’alternative à la conceptualisation des dynamiques de couples. Cette analyse théorique montre ensuite la pertinence d’enrichir les modèles de compatibilité en y introduisant les notions d’attachement et de névrotisme. L’examen de ce modèle est suivi par une description des enjeux méthodologiques et statistiques à prendre en compte pour déterminer les fondements scientifiques des hypothèses sur la compatibilité conjugale. L’analyse de la validité prévisionnelle du modèle de compatibilité proposé compose le cœur de cette thèse. Pour ce faire, des couples effectuant une thérapie conjugale ont complété une série de questionnaires en début de suivi, dont l’Échelle d’ajustement dyadique (Spanier, 1976), le Questionnaire sur les expériences d’attachement amoureux (Brennan, Clark, & Shaver, 1998) et le NEO-FFI (Costa & McCrae, 1992). Les données des 365 couples de l’échantillon sont d’abord analysées en appliquant une méthode fondée sur les modèles d’interdépendance acteur-partenaire (APIM) (Kashy & Snyder, 1995). Les résultats démontrent que la présence d’évitement de la proximité est négativement liée à la satisfaction des conjoints eux-mêmes et à celle de leur partenaire. La méthode APIM révèle aussi que plus l’écart entre les conjoints sur l’évitement et le névrotisme est grand, plus les partenaires rapportent être insatisfaits de leur relation. Les résultats sont ensuite examinés à l’aide d’une technique statistique, l’analyse des regroupements naturels ou des clusters, susceptible de mieux examiner la notion de compatibilité. Cette analyse montre la présence de sept regroupements naturels de conjoints selon leurs scores d’attachement et de névrotisme. De plus, lorsque ces clusters sont croisés en fonction de la détresse conjugale, trois grands sous-types de couples qui engendrent des profils de détresse conjugale distincts émergent des analyses. Ces clusters présentent enfin des différences qualitatives quant à la nature des mandats thérapeutiques confiés aux thérapeutes. Les implications des résultats sont discutées en regard de la validation du modèle de compatibilité fondé sur les notions d’attachement et de névrotisme. Certaines données de la présente thèse suggèrent que ce ne sont pas les caractéristiques de l’un ou de l’autre des conjoints prises isolément, mais plutôt la combinaison des caractéristiques des conjoints qui explique le mieux la nature et la gravité de la détresse conjugale vécue. Plusieurs défis conceptuels, théoriques et méthodologiques sont discutés et des pistes de recherches futures sont proposées afin de consolider les assises scientifiques du modèle de compatibilité. / The first goal of this thesis is to propose a new model of compatibility. Partner’s compatibility is conceptualized from two attachment dimensions (avoidance and anxiety) as well as personality trait (neurotism). The second goal was to test this model on a sample of couples seeking marital therapy. A critical literature review on the role of compatibility in the understanding of couples’ dynamics is completed, underlining the relevance of improving this concept’s definition by taking attachment and neurotism into account. Methodological and statistical issues to consider are also discussed. To evaluate the predictive validity of the model, 365 couples completed the Dyadic adjustment scale (Spanier, 1976), the NEO-FFI (Costa & McCrae, 1992) and the Experiences in close relationships questionnaire (Brennan, Clark, & Shaver, 1998) at intake. The association between personality, attachment and marital satisfaction is first examined with the Actor-Partner Interdependence Model (Kashy & Snyder, 1995). Results show that attachment avoidance is negatively associated with marital satisfaction for spouses themselves as well as for their partners. Analysis of similarities and differences is then performed, revealing that differences between partners on the combination of avoidance and neurotism are correlated with marital insatisfaction. To examine the patterns of natural groupings between partners in the sample, clusters analyses were conducted. Seven different types of couples exhibiting significant differences in their marital satisfaction were found. Three main groups of couples were identified within these types, the first one regrouping mostly satisfied couples, a second one including very dissatisfied partners and a last one presenting a difference between the man’s and the women’s satisfaction. These results could be considered as provinding preliminary confirmation of our compatibility model based on attachment and neurotism, but these propositions must be validated. Some elements of the present findings suggest that it is the combination of attachment anxiety, avoidance and neurotism between partners that is linked to severity of marital distress. However, many challenges, conceptual, theoretical, as well as methodological, are then discussed and future scientific contributions are proposed in order to consolidate the conceptual base of compatibility.

This diploma thesis deals with alternative estimation possibilities of the gravity model in trade. We provide the reader with a synthetic methodological overview of the technical problems with the estimation of gravity equations. Consequently, we test for the heterogeneity of data sets used in gravity models of trade which leads us to a conclusion that behavioural patterns of exporters and importers built in the datasets are very complicated and a single generalized specification of gravity equation can lead to bias in estimates and/or to similarly generalized conclusions that hide important robust idiosyncrasies in behavior present in some subsamples of economic agents. Both the theory of estimation techniques and dataset heterogeneity are applied in the empirical part estimating Austria's export function.

Almost two thousand years ago, Claudius Ptolemy created a guide to drawing maps of the world, identifying the names and coordinates of over 8,000 settlements and geographical features. Using the coordinates of those cities and landmarks which have been identified with modern locations, a series of best-fit transformations has been applied to several of Ptolemy’s regional maps, those of Britain, Spain, and Italy. The transformations relate Ptolemy’s coordinates to their modern equivalents by rotation and skewed scaling. These reflect the types of error that appear in Ptolemy’s data, namely those of distance and orientation. The mathematical techniques involved in this process are all modern. However, these techniques have been altered in order to deal with the historical difficulties of Ptolemy’s maps. To think of Ptolemy’s data as similar to that collected from a modern random sampling of a population and to apply unbiased statistical methods to it would be erroneous. Ptolemy’s data is biased, and the nature of that bias is going to be informed by the history of the data. Using such methods as cluster analysis, Procrustes analysis, and multidimensional scaling, we aimed to assess numerically the accuracy of Ptolemy’s maps. We also investigated the nature of the errors in the data and whether or not these could be linked to historical developments in the areas mapped.

Föreliggande uppsats avser belysa relationen mellan stress och personlighet, detta undersöktes ur ett individorienterat perspektiv som baserades på personlighetsprofiler. Data baserades på en enkätundersökning som mätte personlighet utifrån femfaktormodellen och självupplevd stress. 89 medarbetare på ett svenskt företag deltog i studien, 43 män och 46 kvinnor mellan åldrarna 22 och 66 år. För att klassificera olika personlighetsprofiler genomfördes en klusteranalys. För att sedan undersöka relationen mellan personlighetsprofilerna och upplevd stress jämfördes graden av upplevd stress mellan profilerna. Resultatet visade på att det finns skillnader mellan personlighetsprofilerna avseende graden av upplevd stress. Individer med lägst värden inom extraversion och något lägre värden inom öppenhet i förhållande till de andra dragen och personlighetsprofilerna upplevde högre grad stress. Individer som tillhör personlighetsprofilen med högst värden inom samtliga dimensioner i femfaktormodellen upplevde låg stress. / This paper is intended to describe the relationship between stress and personality, this is examined from an individual-oriented perspective based on personality profiles. The data was based on a questionnaire that measured personality based on the five factor model and self-perceived stress. 89 employees of a Swedish company participated in the study, 43 men and 46 women between the ages of 22 and 66 years. A cluster analysis was used to classify different personality profiles. To later examine the relationship between personality profiles and perceived stress the degree of perceived stress was compared between the profiles. The result showed that there are differences between the personality profiles regarding the level of perceived stress. Individuals with the lowest values in extraversion and slightly lower values of openness in relation to other features and personality profiles, experienced higher levels of stress. Individuals belonging to the personality profile with the highest values in all dimensions of the five factor model experienced low levels of stress.

The first step of the hierarchical approach to the international market segmentation is described and demonstrated in this master's thesis. The theoretical part that covers the methodology, statistical background and secondary data sources is followed by the illustration of cluster analysis. The outcome is the division of thirty European countries into five segments with respect to the post-productive population. The formed segments are then characterised and interpreted. Their usability is shown in three challenges in the European Year of 2012, which is dedicated to Active Ageing.

This thesis focuses on the topic of VAT in EU member states. The aim is to specify groups of member states in accordance to the significance of VAT in national tax systems and to identify factors, which may affect the grouping process. The hierarchical method of cluster analysis based on the 1995, 2002, 2009 and 2017 data is used for sorting states into similarity-based groups. Macroeconomic indicators characterizing VAT are chosen for cluster analysis: VAT per capita, VAT share of GDP, VAT share of total tax revenues, Standard rate and Average reduced rate. Development of cluster structure during the period discovers an increasing number of states separated into one-member clusters for their remarkable dissimilarity. It also indicates, that member states classification as OMS or NMS still remains a significant factor of the clustering process. So-called old and new member states seem to appear together in one cluster rarely. Testing of hypothesis of VAT convergence concerning 28 EU member states during 1995-2017 (in the field of chosen variables) is another sub-target of this paper. Via application of basic statistics on timeline data, and especially evaluation of coefficient of variation development, the hypothesis of VAT convergence in the EU during years 1995-2017 is proven.

This thesis deals with the consumption taxes changes in the Member States of EU in relation with the harmonization process. Its goal is to find out via chosen indicators, whether the influence of the harmonization processes in an area of value added taxes and excise duty, brings Member States of EU closer in an area of consumption taxation. As the main method of data processing is used the cluster analysis for which three indicators of tax burden are selected. Via analysis the cluster processes are used in selected years 2000 and 2017. All Member States are included with the exception of Croatia and that is because absence of the data. The data that are used in the cluster analysis are extracted from the Eurostat and from the database of European Commission. The states are divided according to the similarity of taxation into particular clusters. The results are presented in a form of graphical outputs the so called dendrograms. The conclusion is comparison of researched years according to the harmonization process. According to the cluster analysis results gained via selected consumption tax indicators, there is no converge occurring in 2000 to 2017 between original EU-15 states and the new Member States which have joined EU in 2004 or later.

This research investigates eco-clusters as driver for greening regional economic policy and examines necessary incentive structures to foster eco-innovation as well as growth and employment in the eco-industry sector. Eco-clusters are seen in context with sustainability and environmental friendly behavior as means for a socio-ecological transition in the long run. The main hypothesis implies that eco-clusters have to be policy driven and established top-down and therefore differ from cluster structures in other industries. Possible reasons are uncertainty on a developing market as well as external effects of eco-innovations; the latter are seen as radical innovations. Based on theoretic findings for the establishment of clusters and general research findings for eco-clusters and eco-innovations, it is differentiated between a spontaneous cluster emergence from private initiatives through self-reinforcing forces of companies in a region (bottom-up), and the formation of a policy-driven network with primarily regional objectives to stimulate the competitive advantage of the regional industrial location (top-down). The hypothesis will be proofed by empirical results gained through personal interviews and complemented by findings in current research literature. Finally, implications for incentive structures to green economic policy are identified. It is shown that eco-clusters are different to other clusters and crucial for a long-term sustainable change and thus need political commitment and public incentives. For empirical observation, eco-clusters in Austria were selected. This research relates to the ongoing debate on green growth and develops policy incentives for establishment of eco-clusters and thus greening of economic policy. / Series: WWWforEurope