Statistical analysis of networks has grown rapidly over the last few years with increasing number of applications. Graph-valued data carries additional information of dependencies which opens the possibility of modeling highly complex objects in vast number of fields such as biology (e.g. brain networks , fungi networks, genes co-expression), chemistry (e.g. molecules fingerprints), psychology (e.g. social networks) and many others (e.g. citation networks, word co-occurrences, financial systems, anomaly detection). While the inclusion of graph structure in the analysis can further help inference, simple statistical tasks in a network is very complex. For instance, the assumption of exchangeability of the nodes or the edges is quite strong, and it brings issues such as sparsity, size bias and poor characterization of the generative process of the data. Solutions to these issues include adding specific constraints and assumptions on the data generation process. In this work, we approach this problem by assuming graphs are globally sparse but locally dense, which allows exchangeability assumption to hold in local regions of the graph. We consider problems with two types of locality structure: block structure (also framed as multiple graphs or population of networks) and unstructured sparsity which can be seen as missing data. For the former, we developed a hypothesis testing framework for weighted aligned graphs; and a spectral clustering method for community detection on population of non-aligned networks. For the latter, we derive an efficient spectral clustering approach to learn the parameters of the zero inflated stochastic blockmodel. Overall, we found that incorporating multiple local dense structures leads to a more precise and powerful local and global inference. This result indicates that this general modeling scheme allows for exchangeability assumption on the edges to hold while generating more realistic graphs. We give theoretical conditions for our proposed algorithms, and we evaluate them on synthetic and real-world datasets, we show our models are able to outperform the baselines on a number of settings. <br>

Denna uppsats undersöker huruvida tillväxtföretag genererar en abnormal avkastning i samband med annonsering av ett förvärv. Studien genomför en händelsestudie där 30 noterade företag på Stockholmsbörsen har blivit utvalda genom olika urvalskriterier, till exempel köpeskilling och typ av företag. Vad som klassificerar ett tillväxtföretag kan skilja sig, men denna uppsats har med grund i teorin valt ut två olika finansiella nyckeltal, price-to-book och price-to-sales. Nyckeltalen för de enskilda företagen skall vara större eller minst lika med medianen på den tillväxtorienterade aktielistan Spotlight. Teorier som den effektiva marknadshypotesen, hybrishypotesen samt teorier om varför företag genomför förvärv tas upp och diskuteras. Uppsatsen diskuterar även informationsasymmetri och informationsläckage.Studiens resultat visar att urvalet vid annonseringsdagen genererar en genomsnittlig abnormal avkastning som är positiv, både när marknadsmodellen och constant mean return-modellen används. Urvalet innehåller uteliggare, därav kompletteras medelvärdet med medianen. Resultat för medianen påvisar en positiv abnormal avkastning. Resultatet pekar på att marknaden i genomsnitt är optimistiska till att tillväxtföretag genomför förvärv på kort sikt. / The study examines whether growth companies generate an abnormal return in relation to an announcement of an acquisition. The study conducts an event study where 30 listed companies on the Stockholm Stock Exchange have been selected through various selection criteria, such as purchase price and type of company. How one should classify a growth company differs, but based on the theory, this study has selected two financial key ratios, price-to-book and price-to-sales. The key ratios for the individual company must be greater than or at least equal to the median on the growth-oriented share list Spotlight.Theories such as the efficient market hypothesis, the hubris hypothesis, and theories about why companies make acquisitions are discussed. The study also discusses information asymmetry and leakage of information.The results of the study show that the sample at the day of announcement generates a positive average abnormal return both when the market model and when the constant mean return model are used. The study was affected by outliers, and which the mean was supplemented by the median, where the result showed a positive abnormal return. The results also indicate that the market is on average optimistic to growth companies making acquisitions in short term.

No description available.

Obesity prevalence continues to increase worldwide, yet few safe and effective treatment options are available suggesting there needs to be a greater emphasis on preventing rather than treating obesity. This research investigated the association of obesity predisposing SNPs and a gene score with nutrient consumption patterns including total energy intake and macronutrient distribution in a European ancestry population as well as discussing an evolutionary perspective on eating disorders using current epidemiological evidence to identify genes which may be involved. The association of two of the 14 obesity predisposing SNPs and the gene score with BMI was confirmed in the EpiDREAM population. Novel associations between two SNPs located in or near BDNF (rs6265 and rs1401635) were found with total fat, MUFA, and PUFA intake. Rs1401635 was also associated with total energy and trans fat intake. Novel associations of rs6235 (PCSK1) and the gene score were found with total energy intake. The novel associations found indicate that food related behaviours are one of the mechanisms of action through which obesity predisposing SNPs cause obesity and therefore warrant further investigation. The lack of association among all genes and the modest association of the gene score show that mechanisms other than food consumption are important. The investigation of the evolutionary history of eating disorders revealed that the adapted to flee famine hypothesis is a plausible theory explaining anorexia nervosa while the thrifty genotype hypothesis provides a possible explanation for bulimia nervosa and binge eating disorder. These evolutionary theories can be applied to identify new candidate genes as well as phenotypic traits to investigate to better understand the genetic architecture of eating disorders. Understanding genes associated with disordered eating patterns may highlight future areas for obesity prevention. / Thesis / Master of Science (MSc) / A large percentage of the risk of developing obesity or an eating disorder (anorexia nervosa, bulimia nervosa, and binge eating disorder) is determined by genetics. For obesity, many genes have been identified as influencing risk, but the mechanisms through which the genes work are largely unknown. For eating disorders, gene identification efforts have been mostly unsuccessful and no mechanisms of action have been determined. In the first component of this thesis we found an association between previously identified obesity risk genes and food intake, specifically the total number of calories consumed per day and the percentage of calories from total fat and fat subtypes. These results support that food related behaviours are possible mechanisms of action which need to be further investigated. In the second half of the thesis we viewed eating disorder behaviours from an evolutionary perspective. We concluded that there are theories that possibly explain eating disorder behaviours including being able to live off of small quantities of food as well as binging. These evolutionary theories can be applied to identify new genes to study in the context of eating disorders as well as different definitions of eating disorders.

In this thesis we provide a body of knowledge that concerns Riemann zeta-function and its generalizations in a cohesive manner. In particular, we have studied and mentioned some recent results regarding Hurwitz and Lerch functions, as well as Dirichlet's L-function. We have also investigated some fundamental concepts related to these functions and their universality properties. In addition, we also discuss different formulations and approaches to the proof of the Prime Number Theorem and the Riemann Hypothesis. These two topics constitute the main theme of this thesis. For the Prime Number Theorem, we provide a thorough discussion that compares and contrasts Norbert Wiener's proof with that of Newman's short proof. We have also related them to Hadamard's and de la Vallee Poussin's original proofs written in 1896. As far as the Riemann Hypothesis is concerned, we discuss some recent results related to equivalent formulations of the Riemann Hypothesis as well as the Generalized Riemann Hypothesis.

Doctor of Philosophy / Department of Statistics / Haiyan Wang / Statistical tools to detect nonlinear relationship between variables are commonly needed in various practices. The first part of the dissertation presents a test of independence between a response variable, either discrete or continuous, and a continuous covariate after adjusting for heteroscedastic treatment effects. The method first involves augmenting each pair of the data for all treatments with a fixed number of nearest neighbors as pseudo-replicates. A test statistic is then constructed by taking the difference of two quadratic forms. Using such differences eliminate the need to estimate any nonlinear regression function, reducing the computational time. Although using a fixed number of nearest neighbors poses significant difficulty in the inference compared to when the number of nearest neighbors goes to infinity, the parametric standardizing rate is obtained for the asymptotic distribution of the proposed test statistics. Numerical studies show that the new test procedure maintains the intended type I error rate and has robust power to detect nonlinear dependency in the presence of outliers. The second part of the dissertation discusses the theory and numerical studies for testing the nonparametric effects of no covariate-treatment interaction and no main covariate based on the decomposition of the conditional mean of regression function that is potentially nonlinear. A similar test was discussed in Wang and Akritas (2006) for the effects defined through the decomposition of the conditional distribution function, but with the number of pseudo-replicates going to infinity. Consequently, their test statistics have slow convergence rates and computational speeds. Both test limitations are overcome using new model and tests. The last part of the dissertation develops theory and numerical studies to test for no covariate-treatment interaction, no simple covariate and no main covariate effects for cases when the number of factor levels and the number of covariate values are large.

Master of Science / Department of Statistics / Leigh W. Murray / The term retrospective power describes methods for estimating the true power of a test after data have been collected. These methods have been recommended by some authors when null hypothesis of a test cannot be rejected. This report uses simulations to study power as a construct of an observed effect, variance, sample size, and set level of significance under the balanced one-way analysis of variance model for normally distributed populations with constant variance. Retrospective power, as a construct of sample data, is not recommended when the null hypothesis of a test cannot be rejected. When the p-value of the test is large, estimates for true power tend to fall below the 0.80 level and width-minimized confidence limits for true power tend to be wide.

Opposition to Paul as central to the occasion of the letter to the Galatians ·(expressed in scholarly circles in terms of different opponent hypotheses) is identified as a pivotal factor in the interpretation of that letter. An analysis of some of the received opponent hypotheses reveals that today, as throughout the history of the critical interpretation of the letter, they are basically different proposals sharing the same historical and methodological components. Both components are critically examined and it is suggested that inadequate assumptions lead to an anachronistic and ethnocentric view of the nature of the conflict and consequently, from a his­ torical point of view, a distorted cypher key to the interpretation of the letter. The proposal of a social-scientific approach to the letter presents, at the meth­ odological level, the components of an alternative cypher key. It attempts to be interdisciplinary in that it accounts for the otherness of foreign cultures and distant historical eras in an integrated and explicit way. Because of the methodological components an alternative set of historical com­ ponents can be suggested. A focus on contemporary views on first-century Judaism and the expansion of the Pauline communities in the first-century world not only . points to shortcomings in the received views but provides an alternative perspective on the communicative context of the letter. A first but important step is taken to construct a probable first-century Mediterranean communicative context for the interpretation of the letter. In view of this alternative communicative context it is possible to redefine the con­ flict as a truly first-century Mediterranean one in which Paul's authority and honour were at stake. This makes it possible to construe the text in a different way. Instead of regarding the letter to the Galatians as a document of intra-Christian struggle about conflicting theological or doctrinal convictions, it is suggested that the letter be construed as the product of Paul's missionary endeavour within the household setting, the subculture of first-century Judaism, and the confines of orality and reli­ gion in the first-century world. / Biblical and Ancient Studies / D. Th. (New Testament)

This paper researches the effects of choice of payment (cash and stock) and M&amp;A type (technological and non-technological) on bidders’ performance. We investigate 500 events in Swedish market between 2005 and 2015. Moreover, we also control the size of firms and the value of takeovers. In this paper, we conduct an event study in order to generate abnormal returns for the bidders at and around the M&amp;A announcement. This research generates statistically significant and positive abnormal returns for the bidders especially when deals are financed by shares. Moreover, we also find that the technological M&amp;A brings about lower abnormal returns than non-technological M&amp;A. When we control for payment choice in technological M&amp;A, the result shows that technological M&amp;A paid for in shares generates higher abnormal returns than technological M&amp;A paid for in cash.

Thesis (MSc)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: See the full text for the abstract / AFRIKAANSE OPSOMMING: Sien volteks vir die opsomming