Este trabalho analisou o cenário em que está inserida a comercialização de bovinos para abate, no estado do Rio Grande do Sul. Para isto, foram utilizados dados obtidos através das Guias de Trânsito Animal (GTAs), entre os anos de 2010 (ano em que o sistema informatizado para GTAs foi implantado na Secretaria da Agricultura, Pecuária e Agronegócio - SEAPA) a 2013. Também foram utilizados dados da Declaração Anual de Rebanho, fornecidos pela SEAPA. Um dos diferenciais deste trabalho é esta fonte única, com documentos obrigatórios a todos que comercializam animais e o agrupamento por mesorregião geográfica das informações obtidas. Para o entendimento deste cenário, foram utilizadas como aporte teórico a Teoria da Localização, que considera a otimização da localização de uma firma, e a Organização Industrial, que através do Modelo E-C-D proporciona subsídios para caracterização da concentração em uma indústria. Os resultados mostraram que o número de animais abatidos é diferente entre as mesorregiões geográficas. Os matadouros-frigoríficos recebem animais para abate provenientes da mesorregião geográfica em que se estão localizados, mas também de outras mesorregiões do RS, em diferentes proporções. Nas análises, as mesorregiões geográficas Centro Oriental e Metropolitana destacaram-se pelo maior número de animais abatidos. No entanto, o rebanho nestas mesorregiões ainda não foi o suficiente para a demanda gerada pelo mercado consumidor, que está muito próximo devido as grandes cidades que compõem estas mesorregiões, como Santa Maria e Porto Alegre. A Mesorregião Sudoeste é a terceira que mais abateu animais. A quantidade abatida nesta mesorregião é composta por 74% de animais da própria mesorregião Sudoeste, porém, o seu rebanho foi uma importante fonte de animais para as demais mesorregiões. Neste cenário, o estado do RS apresentou mais de 60 mil ofertantes (pecuaristas) em 2013, e o número de demandantes (matadouros-frigoríficos) chegou a 459. O matadouro-frigorífico que mais abateu em 2010 perdeu participação de mercado até 2013, e frigoríficos menores conseguiram aumentar as suas participações, havendo uma tendência de maior equilíbrio entre as empresas. Não existe concentração na indústria de carne bovina, e as medidas de concentração analisadas mostram índices cada vez menores. / This study analyzed the scenario in which the marketing of cattle for slaughter is inserted in the state of Rio Grande do Sul. For this, it was used data of Animal Transit Guides (GTAs) for slaughter since 2010 - year that these guides were made available electronically by the Secretariat of Agriculture, Livestock and Agribusiness (SEAPA) - to 2013. It was also used data from Herd Annual Statement, supplied by SEAPA, which is one of the remarkable points of this paper, the unique and reliable source that was provided with a set of mandatory documents that all cattle sellers must have. Based on information that was founded, it was used the theory of location, considering the optimization of the location of a firm, and the Industrial Organization, through the S-C-P model, focusing on the structure that allows one to see if there was concentration in the industry. The Metropolitan ant Eastern Center Mesoregions are the ones that slaughter the most, having their own production as their raw material. However, these cattle are still not enough for the demand generated by the consumer market, which live close to these centers. The Southwest Mesoregion slaughters a great number of animals, with 74% of these slaughtered cattle coming from their own production and it is still an important source of animals for other Mesoregions. The number of suppliers (farmers) exceeded 60,000 in 2013, and the slaughter houses had reached 459. There is no concentration in the beef industry, and analyzed concentration measurements show dwindling numbers. The most important slaughterhouse in 2010 lost market by 2013, and smaller slaughterhouses were able to increase their holdings, with a trend towards greater balance between companies.

L’externalisation des données dans le Cloud a engendré de nouvelles problématiques de sécurité. L’enjeu est de protéger les données des utilisateurs et leur vie privée. En ce sens, deux principes ont été suivis durant cette thèse : le premier est d’avoir une confiance limitée envers l’hébergeur de données (entre autres), le deuxième est d’établir une architecture basée sur un modulede confiance placé en rupture entre le poste client et le Cloud, d’où l’approche "Trust The Module,Not The Cloud" (TTM).Déléguer donc les opérations de sécurité à un module matériel dédié permet alors plusieurs bénéfices : d’abord s’affranchir d’un poste client davantage vulnérable face à des attaques internes ou externes ; ensuite limiter les composants logiciels au strict minimum afin d’avoir un meilleur contrôle du fonctionnement et enfin dédier les opérations cryptographiques à des co-processeurs spécialisés afin d’obtenir des performances élevées. Ainsi, les travaux menés durant cette présente thèse suivent trois axes. Dans un premieraxe nous avons étudié les défis d’un Cloud personnel destiné à protéger les données d’un particulier, et basé sur une carte nano-ordinateur du marché peu coûteuse. L’architecture que nous avons définie repose sur deux piliers : une gestion transparente du chiffrement grâce à l’usage d’un chiffrement par conteneur appelé Full Disk Encryption (FDE), initialement utilisédans un contexte de protection locale (chiffrement du disque d’un ordinateur ou d’un disque dur externe) ; et une gestion transparente de la distribution grâce à l’usage du protocole iSCSI qui permet de déporter le conteneur sur le Cloud. Nous avons montré que ces deux piliers permettent de construire un service sécurisé et fonctionnellement riche grâce à l’ajout progressif de modules"sur étagère" supplémentaires.Dans un deuxième axe, nous nous sommes intéressés au problème de performance lié à l’usage du FDE. Une étude approfondie du mode de chiffrement XTS-AES recommandé pour le FDE, du module noyau Linux dm-crypt et des co-processeurs cryptographiques (ne supportant pas tous le mode XTS-AES), nous ont conduit à proposer différentes optimisations dont l’approche extReq, qui étend les requêtes cryptographiques envoyées aux co-processeurs. Ces travaux nousont ainsi permis de doubler les débits de chiffrement et déchiffrement.Dans un troisième axe, afin de passer à l’échelle, nous avons utilisé un module de sécurité matériel (Hardware Secure Module ou HSM) certifié et plus puissant, dédié à la protection des données et à la gestion des clés. Tout en capitalisant sur l’architecture initiale, l’ajout du module HSM permet alors de fournir un service de protection adapté aux besoins d’une entreprise par exemple. / Data outsourcing to the Cloud has led to new security threats. The main concerns of this thesis are to protect the user data and privacy. In particular, it follows two principles : to decrease the necessary amount of trust towards the Cloud, and to design an architecture based on a trusted module between the Cloud and the clients. Both principles are derived from a new design approach : "Trust The Module, Not The Cloud ".Gathering all the cryptographic operations in a dedicated module allows several advantages : a liberation from internal and external attacks on client side ; the limitation of software to the essential needs offers a better control of the system ; using co-processors for cryptographic operations leads to higher performance.The thesis work is structured into three main sections. In the first section , we confront challenges of a personal Cloud, designed to protect the users’ data and based on a common and cheap single-board computer. The architecture relies on two main foundations : a transparent encryption scheme based on Full Disk Encryption (FDE), initially used for local encryption (e.g., hard disks), and a transparent distribution method that works through iSCSI network protocol in order to outsource containers in Cloud.In the second section we deal with the performance issue related to FDE. By analysing the XTS-AES mode of encryption, the Linux kernel module dm-crypt and the cryptographic co-processors, we introduce a new approach called extReq which extends the cryptographic requests sent to the co-processors. This optimisation has doubled the encryption and decryption throughput.In the final third section we establish a Cloud for enterprises based on a more powerful and certified Hardware Security Module (HSM) which is dedicated to data encryption and keys protection. Based on the TTM architecture, we added "on-the-shelf" features to provide a solution for enterprise.

Master of Agribusiness / Department of Agricultural Economics / Jason Bergtold / The dramatic changes that have taken place in the production agriculture industry in the last decade have the Long Family Partnership wanting to reassess their farm land management strategy. As land owners, they feel as though they might be missing out on profit opportunity by continuing their current lease agreements as status quo. The objective of this research is to determine the optimal land management strategy for the Partnership farm that maximizes net returns for crop production, but also taking into account input costs and risk. Three scenarios were built: (1) a Base Case of the current share-crop and cash lease Agreements; (2) the possibility of farming their own irrigated farm land and continuing to cash lease land used to produce dryland wheat; and (3) deciding to farm all the irrigated and dry land farm acreage themselves. In order to do this, a whole-farm budget spreadsheet model was generated to assess alternative land management scenarios. The difference in net returns between alternative land rental scenarios were then compared and followed by a sensitivity analysis and stochastic analysis using @RISK software. The findings concluded that there was greater potential to increase net farm income while still conservatively managing risk by investing into their own farm land, as not only owners but also as operators. The stochastic and sensitivity analysis confirmed that farming their own land was more sensitive to changes in yields, prices and input expenses. However, even in consideration of the additional risk, the probability of increasing net farm income was greater for the scenarios in which they farmed their own land.

Master of Agribusiness / Department of Agricultural Economics / Allen M. Featherstone / The Agriculture Equipment Manufacturing industry is a $42 billion dollar industry in the United States. The Agricultural Equipment industry is very competitive across all market segments, especially in the less than 100 horsepower category (<100hp). This tractor category consists of 4 sub categories: <20hp, 20-40hp, 40-60hp, and 60-100hp. The <100hp tractor segment accounted for 170,547 of the 207,833 tractors that were sold during the 2014 year. Compared to the over 100 horsepower category (100+hp) that has fewer competitors, the <100hp segment is more competitive with more manufacturers competing for market share. Company XYZ is a full line manufacturer of agricultural equipment, harvesters, and construction equipment. Company XYZ lost some ground in market share due to the increased competition from new entrants into the market place as well as established manufacturers increasing their presence. To be more competitive, Company XYZ is looking at industry best practices to see how they can increase market share. One of these practices is a terms bank. A terms bank allows a dealer to stockpile unused months of terms to be used at a later date on tractors with expired terms. This minimizes financial risk for dealers to stock inventory. The cost to stock inventory is a large expense that dealers must carefully manage. One of the biggest costs of stocking inventory is the interest paid for tractors that have exhausted their interest free terms. A terms bank may lower the amount of interest that a dealer pays. It also lowers the cost to stock inventory and allows the dealership to manage and reduce these costs and risks. Evaluating the factors associated with stocking inventory, especially interest rate, will help manage inventory costs and stocking levels. This thesis uses regression analyses to analyze the costs of stocking units and the effect it has on dealership revenues. A regression analysis will test the hypothesis that lowering the interest portion of the cost of stocking inventory will increase sales. Data were gathered for dealership groups in the Western United States on a monthly basis for the years 2008 – 2014. The results supported the hypothesis that lowering the interest rate at dealerships was positively correlated with revenues. The reduced interest cost lowers the carrying cost of inventory and point to a terms bank being an effective tool for increasing Company XYZ’s market share.

Master of Science - Agricultural Education and Communication / Department of Communications and Agricultural Education / Lauri M. Baker / The equine industry is an established part of Maryland Agriculture; the most recent equine census placed 79,100 equines valued at approximately $714 million in the state, with approximately ten percent of those animals housed in Montgomery County. But, equine operators are a unique demographic in the agricultural realm. They are not managing their lands to produce food or fiber and often are employed in other professions unrelated to agriculture. Unlike other agricultural operators, they are often unaware of conservation terms or practices and programs available to help implement these practices on their land. The basic objective of this study was to explore the level of understanding equine operators in Montgomery County, Maryland have of conservation best management practices (BMPs), associated cost-share programs, industry regulations and the aspects and methods of communication that influence them in gaining awareness of and the adoption of these practices on their properties. The study applied the tenets of the decomposed theory of planned behavior (DTPB), which combines the theory of planned behavior (TPB), diffusion of innovation theory (DOI) and economic constraint theory gathered through semi-structured qualitative interviews with participants, selected using purposive and snowball sampling. Utilizing the grounded theory method to discover emerging patterns in the data resulting from these interviews aided in identifying the most effective means of educating equine operators regarding BMPs and increasing the adoption of these practices on equine properties within the County. Equine operators in the study were found to use a variety of informational sources, had a high level of adaptation of the BMPs they used and overall saw their role as land stewards as an important aspect of their environmental actions.

abstract: Over the last two decades, Alternative Project Delivery Methods (APDM), such as Design-Build (DB), have become more popular in the construction industry, specifically in the U.S., and the competition for APDM projects has risen among construction companies. The Engineering News Record (ENR) magazine analyzes DB firms and publishes the list of the top 100 every year. According to ENR articles and many scientific papers, the implementation of DB method has grown drastically over the last decade, however, information about growth trends depending on firm size and segment is lacking. Also missing is knowledge the future market trends over the next five years. Furthermore, public agencies and DB firms may be worried that DB projects do not distribute wealth equally among DB firms. Using the top 100 firms deemed representative of the DB market, the author has divided the market into volumes based on rankings to analyze the total DB market revenue growth. A comparison between international and domestic revenues indicated that the top five DB firms have 64% more involvement in the international market compared to the domestic market. Furthermore, while the research shows increasing market share only for the top five firms, the author has found that (1) a large portion of their market share is due to a large growth in their international market, and (2) revenues for all volumes of the DB market have increased. Moreover, regression and time series analyses allow for the forecasting of the DB market growth, which the author anticipate to move from about $100B to about $150B in 2020. / Dissertation/Thesis / Masters Thesis Construction 2014

abstract: This thesis starts with an analysis of the current situation of the pharmaceutical industry in China, and discusses the strategic mergers and acquisitions (M&A) by small and medium-sized enterprises (SMEs) in the pharmaceutical industry in China. It elaborates on the rationale for the development of the mergers and acquisitions of the pharmaceutical SMEs. Then a literature review is provided on the causes of corporate mergers and acquisitions such as the economies of scale, synergistic effect, transaction costs, market power, and strategic diversification.Next,the thesis analyzes the underlying rationale for the M&A transactions in the pharmaceutical industry in China, and explores the likely path of successful value creation for pharmaceutical SMEs in China. Specifically, with five in-depth case studies of M&A transactions of pharmaceutical firms, this thesis reveals the critical success factors leading to value creation and growth in the practice of mergers and acquisitions of the pharmaceutical SMEs in China. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2015

abstract: Mobile data collection (MDC) applications have been growing in the last decade especially in the field of education and research. Although many MDC applications are available, almost all of them are tailor-made for a very specific task in a very specific field (i.e. health, traffic, weather forecasts, …etc.). Since the main users of these apps are researchers, physicians or generally data collectors, it can be extremely challenging for them to make adjustments or modifications to these applications given that they have limited or no technical background in coding. Another common issue with MDC applications is that its functionalities are limited only to data collection and storing. Other functionalities such as data visualizations, data sharing, data synchronization and/or data updating are rarely found in MDC apps. This thesis tries to solve the problems mentioned above by adding the following two enhancements: (a) the ability for data collectors to customize their own applications based on the project they’re working on, (b) and introducing new tools that would help manage the collected data. This will be achieved by creating a Java standalone application where data collectors can use to design their own mobile apps in a userfriendly Graphical User Interface (GUI). Once the app has been completely designed using the Java tool, a new iOS mobile application would be automatically generated based on the user’s input. By using this tool, researchers now are able to create mobile applications that are completely tailored to their needs, in addition to enjoying new features such as visualize and analyze data, synchronize data to the remote database, share data with other data collectors and update existing data. / Dissertation/Thesis / Masters Thesis Computer Science 2016

abstract: Since the 2008 financial crisis, the total assets managed by U.S. mutual funds have frequently hit new highs and the industry has become increasingly concentrated. In the meantime, two strategies have emerged in the American mutual fund industry: active and passive management. What factors affect the market shares of firms that adopted these two different strategies? Building on strategic management theories, I suggest that mutual fund families that adopted active and passive management strategies tend to compete in different dimensions. Active management fund families tend to implement the product differentiation strategy, competing on “product quality” through excess-returns, innovative and differentiated fund products; passively managed fund families focus more on "price competition" by conducting an overall cost leadership strategy. This research examines the driven factors of fund families’ market share. The results show that: the market share of actively managed fund families is more sensitive to positive impact of fund performance, while passive management firms are more sensitive to negative effect of management fees and total loads; 12b-1 expense improves the competitiveness of active fund families and thus enhance their market shares but it has negative impact on passive fund families. In addition, high turnover decreases the market share of all fund families, especially for passively managed families. The outcome reveals the latest US mutual industry orientation: products differentiation, turnover, management fee have greater impact on market share while the competition of fund performance is diminishing. The Matthew effect in US mutual fund industry is outstanding. Industrial competition dimension expands from performance and products to cost cutting. Empirical analysis on Chinese mutual fund families is also conducted. Different from the US, there is only small number of mutual fund families targeting passive management products. The results show that the distribution channel has the largest impact on Chinese mutual fund family market share and investors are more willing to chase performance than to consider cost-efficient fund families. This study then analyses reasons behind the difference of Chinese and American mutual fund industries. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2018

Incineration of waste materials has become a common way to handle a worldwide increase of produced waste materials. The problem with waste as fuel is that the material is not homogeneous and include a mixture of fossils and renewables. The fossil part ends up in an emission of fossil carbon dioxide in a combustion process, which is included in the emissions trading system. However, since waste material varies a lot in composition depending on the time of the year, origins, etc. it is preferable to have a method for real-time measurement of the fossil share of combusted waste. No real-time measurement technologies are available today which is the reason to investigate if near-infrared (NIR) spectroscopy could be a potential solution. An artificial mixture of refused derived fuel has been used to investigate the possibilities of NIR for prediction of the fossil share in waste material. The fossil share is assumed to be equal to the content of plastic material with an origin of oil products. Mixtures with different plastic content are scanned by the NIR instrument to obtain individual absorption spectra. A Partial least square (PLS) regression model is created on measured spectra and known content of plastics. The best model for the prediction on new spectral data using one of four measured replicates is a PLS model preprocessed with Savitzky-Golay smoothing that gives an R-square value of 0,782. If the prediction is done, depending on a delimitated wavenumber interval and an average of all four replicates is the best model instead of a PLS model pre-processed with standard normal variate without seven outliers that have an R-square value of 0,81. R-square value is the coefficient of determination which has been used to figure out the best model. An R-square value above 0,65 are recommended for process modelling, where 1 is the highest possible value.