Certains jeux de données présentent des caractéristiques géométriques et topologiques non triviales qu'il peut être intéressant d'inférer.Cette thèse traite des vitesses non-asymptotiques d'estimation de différentes quantités géométriques associées à une sous-variété M ⊂ RD. Dans chaque cas, on dispose d'un n-échantillon i.i.d. de loi commune P ayant pour support M. On étudie le problème d'estimation de la sous-variété M pour la perte donnée par la distance de Hausdorff, du reach τM, de l'espace tangent TX M et de la seconde forme fondamentale I I MX, pour X ∈ M à la fois déterministe et aléatoire.Les vitesses sont données en fonction la taille $n$ de l'échantillon, de la dimension intrinsèque de M ainsi que de sa régularité.Dans l'analyse, on obtient des résultats de stabilité pour des techniques de reconstruction existantes, une procédure de débruitage ainsi que des résultats sur la géométrie du reach τM. Une extension du lemme d'Assouad est exposée, permettant l'obtention de bornes inférieures minimax dans des cadres singuliers. / Some datasets exhibit non-trivial geometric or topological features that can be interesting to infer.This thesis deals with non-asymptotic rates for various geometric quantities associated with submanifolds M ⊂ RD. In all the settings, we are given an i.i.d. n-sample with common distribution P having support M. We study the optimal rates of estimation of the submanifold M for the loss given by the Hausdorff metric, of the reach τM, of the tangent space TX M and the second fundamental form I I MX, for X ∈ M both deterministic and random.The rates are given in terms of the sample size n, the instrinsic dimension of M, and its smoothness.In the process, we obtain stability results for existing reconstruction techniques, a denoising procedure and results on the geometry of the reach τM. An extension of Assouad's lemma is presented, allowing to derive minimax lower bounds in singular frameworks.

archives@tulane.edu / Modern biomedical studies often collect multiple types of high-dimensional data on a common set of objects. A representative model for the integrative analysis of multiple data types is to decompose each data matrix into a low-rank common-source matrix generated by latent factors shared across all data types, a low-rank distinctive-source matrix corresponding to each data type, and an additive noise matrix. We propose a novel decomposition method, called the decomposition-based generalized canonical correlation analysis, which appropriately defines those matrices by imposing a desirable orthogonality constraint on distinctive latent factors that aims to sufficiently capture the common latent factors. To further delineate the common and distinctive patterns between two data types, we propose another new decomposition method, called the common and distinctive pattern analysis. This method takes into account the common and distinctive information between the coefficient matrices of the common latent factors. We develop consistent estimation approaches for both proposed decompositions under high-dimensional settings, and demonstrate their finite-sample performance via extensive simulations. We illustrate the superiority of proposed methods over the state of the arts by real-world data examples obtained from The Cancer Genome Atlas and Human Connectome Project. / 1 / Zhe Qu

abstract: Evidence is mounting to address and reverse the effects of environmental neglect. Perhaps the greatest evidence for needing environmental stewardship originates from the ever-increasing extreme weather events ranging from the deadly wildfires scorching Greece and California to the extreme heatwaves in Japan. Scientists have concluded that the probability and severity for about two thirds of such extreme natural events that occurred between 2004 and 2018 is contributed by rising global temperatures. Operations management literature regarding environmental issues have typically focused on the “win-win” approach with a multitude of papers investigating a link between sustainability and firm performance. This dissertation seeks to take a different approach by investigating firm responses to climate change. The first two essays explore firm emissions goals and the last essay investigates firm emissions performance. The first essay identifies firm determinants of greenhouse gas (GHG) reduction targets. The essay leverages Behavioral Theory of the Firm (BTOF) and argues for two additional determinants, Data Stratification and Science-Based Targets, unique to GHG emissions. Utilizing system generalized method of moments on a dataset from Carbon Disclosure Project for years 2011-2017, the paper finds partial confirmation for BTOF and support for the two additional determinants of firm GHG emission goals. The second essay is an exploratory study that seeks to understand factors for firm participation in the Science-Based Targets (SBT) initiative by combining both primary and secondary data analysis. The study is a working paper with primary data still needing to be completed. Secondary data analysis begins with a review of the literature which suggested four potential factors: ISO 14001 certification, Customer Engagement, Emission Credit Purchases, and presence of Absolute Emissions Targets. Preliminary results using panel logistic regression suggest that Emissions Credit Purchases and Absolute Emissions Targets influence SBT participation. The third essay seeks to understand whether stakeholder pressure drives firm GHG emissions reductions. This relies on Stakeholder Theory and classification schemes proposed in Management literature to divide stakeholders, based on their relationship with the firm, into three groups: primary, secondary, and public. Random effects estimation results provide evidence for primary and public stakeholder pressure impacting firm GHG emissions. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2020

The dissertation seeks to meld the two dominant competing theories of party identifi?cation in the US context: the expressive view, where Party ID is seen as a long standing psychological attachment to a political party; and the instrumental view, where Party ID is subject to reevaluation. Using ANES panel data, the paper examines both expressive and instrumental elements of partisanship. In keeping with past research, it finds strong evidence for the expressive understanding of Party ID; partisan groupings tend to be highly stable. However, the strength of identifications varies considerably over time, with perceptions of candidates, presidential approval, policy preferences, and ideological orientations driving these changes. These results are in keeping with an instrumental conceptualization of partisan identities.

The thesis analyses the customer behavior of Kiwi.com, a global online retail company for booking flights and proposes marketing interventions to increase the conversion rates in various customer segments. The effectiveness of new behavioral interventions is tested against current marketing efforts using experimental A/B methods. Additionally, areas for further improvements are explored and a design of future product features and marketing behavioral interventions is proposed.

In this thesis, we describe an approach that won a psychiatric symptom severity prediction challenge. The challenge was to correctly predict the severity of psychiatric symptoms on a 4-point scale. Our winning submission uses a novel stacked machine learning architecture in which (i) a base data ingestion/cleaning step was followed by the (ii) derivation of a base set of features defined using text analytics, after which (iii) association rule learning was used in a novel way to generate new features, followed by a (iv) feature selection step to eliminate irrelevant features, followed by a (v) classifier training algorithm in which a total of 22 classifiers including new classifier variants of AdaBoost and RandomForest were trained on seven different data views, and (vi) finally an ensemble learning step, in which ensembles of best learners were used to improve on the accuracy of individual learners. All of this was tested via standard 10-fold cross-validation on training data provided by the N-GRID challenge organizers, of which the three best ensembles were selected for submission to N-GRID's blind testing. The best of our submitted solutions garnered an overall final score of 0.863 according to the organizer's measure. All 3 of our submissions placed within the top 10 out of the 65 total submissions. The challenge constituted Track 2 of the 2016 Centers of Excellence in Genomic Science (CEGS) Neuropsychiatric Genome-Scale and RDOC Individualized Domains (N-GRID) Shared Task in Clinical Natural Language Processing.

Le satellite spatial Kepler a détecté des oscillations de type solaire parmi plusieurs centaines d'étoiles, permettant la détermination de leurs propriétés physiques à l'aide de l’astérosismologie. Les modèles d'évolution stellaire et les lois d'échelle employés pour déterminer les paramètres tels que la masse, le rayon et l'âge nécessitent toutefois une calibration adaptée. Dans ce contexte, l'utilisation des systèmes binaires présentant des oscillations de type solaires pour les deux étoiles semble particulièrement appropriée. Au cours de cette thèse, nous avons procédé à un travail de détection de ces binaires sismiques parmi les données de Kepler ainsi qu'au développement des outils nécessaires à leur analyse. Bien que la découverte d'une nouvelle binaire sismique semblait très peu probable, nous avons pu rapporter pour la toute première fois la détection d'oscillations de type solaire associées aux deux étoiles les plus brillantes d'un système triple, à savoir HD 188753. À partir de la modélisation, nous avons déterminé des âges semblables pour les deux étoiles détectées en astérosismologie, comme attendu en raison de leur origine commune. Par ailleurs, nous avons entrepris la première analyse orbitale de ce système hiérarchique dans le but d'obtenir une estimation directe des masses et de la parallaxe. Finalement, l'exemple de HD 188753 illustre notre capacité à détecter et à modéliser chacune des étoiles d'un système binaire ou multiple tout en réalisant l'analyse orbitale de celui-ci. Les différents outils développés au cours de cette thèse seront intensivement utilisés dans le cadre des futures missions TESS et PLATO. / The Kepler space telescope detected solar-like oscillations in several hundreds of stars, providing a way to determine their physical properties using asteroseismology. However, the stellar evolutionary models and scaling relations employed to determine parameters such as the mass, the radius and the age require a proper calibration. In this context, the use of seismic binaries showing solar-like oscillations in both stars is especially suitable. During this thesis, we have worked towards the detection of such seismic binaries from the Kepler database and developed the necessary tools to study them. Although the discovery of a new seismic binary was very unlikely, we were able to report for the first time the detection of solar-like oscillations in the two brightest stars of a triple stellar system, namely HD 188753. Using stellar modelling, we found compatible ages for the two stars derived from asteroseismology, as expected from their common origin. In addition, we performed the first orbital analysis of this hierarchical system in order to derive a direct estimate of masses and parallax. Finally, the example of HD 188753 shows our capability to detect and model each of the stars of a binary or multiple system and to perform the orbital analysis of this one. The various tools developed during this thesis will be extensively used in the context of the future missions TESS and PLATO.

The purpose of the study was to determine how South African mathematics teachers were integrating computers into their classrooms. The study was a response to the low achievement scores in mathematics as attained by grade nine learners in the 2011 Trends in International Mathematics and Science Study (TIMSS). TIMSS 2011 assessed Grade four and eight learners. However, South Africa as well as Botswana and Honduras opted to administer the Grade eight assessment to their Grade nine learners instead. South Africa’s Grade nine learners achieved an average score of 352 (35.2%) out of a possible 1 000 points. This quantitative secondary data analysis study utilised data collected from mathematics teachers from 298 schools in South Africa. The dataset was analysed using descriptive analysis that included percentages as well as the Pearson two-way Chi-square tabulations. The major finding of the study is that 73. 9% of South African mathematics teachers are still not integrating computers into mathematics education. Results showed that teachers are mostly using computers for preparation (35.5%) and administration (65.3%) purposes. Even though 45.5% of the teachers reported that they feel comfortable using computers, others feel that they are still in need of technical support. Moreover, the findings showed that 64.8% of the teachers do not attend professional development programmes that focus on the integration of Information Technology (IT) into mathematics. / Dissertation (MEd)--University of Pretoria, 2017. / Science, Mathematics and Technology Education / MEd / Unrestricted

of the Ph.D. Thesis Netrdová, P.: Current trends in quantitative analysis of geographical data: potentialities and limitations of spatial analysis The thesis is a contribution to the discussion about the potentialities of the quantitative approach in geography. It follows the current trends in quantitative analysis of geographical data, specifically spatial analysis, particularly from the perspective of changes in the concept and character of applied methods and their possible contribution in geographical research. Due to the research focus of the author, the entire work is focused primarily on the issue of using quantitative methods in terms of social geography. Attention is focused particularly on statistically spatial analyses, which are the most widely used techniques in social geography, with a wide range of possible applications. One of the goals of this work is to bring the current development in quantitative geography closer to the Czech academic community, and thus contribute to the increased awareness of the potentialities of the application of quantitative methods and spatial analyses in geographical research. Methodological problems in the analysis of spatial data, theoretical changes in the concept of quantitative analysis and also newly emerging quantitative methods have not so far...

In the thesis we study the decay of B0 meson to D∗ s and ρ mesons. The thesis explains the methods and approaches to data analysis in so-called B-factories, simmilar to the KEKB accelerator. The aim of this thesis is to calculate the branching fraction of this decay to further improve the previously measured branching ratio at BaBar experiment with additional data gained from an ex- periment with higher integrated luminosity. This thesis's prospect is not only broadening our knowledge of branching fractions of B0 meson decays, but also it is a starting point for further analysis with the goal of broadening our knowl- edge of CP symmetry violation in the Standard Model by measuring angles of the Unitary Triangle. 1