Bakgrund och problemdiskussion: Företag motiverar vanligtvis centraliserad finansiering med att det leder till värdeökande effekter, något som forskningsvärlden endast teoretiskt berört vilket ger indikationer på att en centraliserad finansiering kan utgöra en värdeökande strategi. Det föreligger därmed ett intresse att undersöka detta empiriskt. Avsaknaden av ett mått för centraliseringsgraden av finansiering medför att framställande av detta mått dock måste genomföras före teorin om värdeökning kan testas empiriskt. Syfte: Studien syftar till att 1) ta fram ett mått för centraliseringsgraden av finansiering, 2) säkerställa validiteten i detta mått och 3) genom detta mått testa centraliseringsgradens effekt på företagsvärde. Metod: Första syftet uppfylls genom en explorativ studie där närliggande teorier använts tillsammans med observationer av verkligheten. Andra syftet uppfylls genom en mindre statistisk undersökning utifrån årsredovisningar samt en enkätundersökning riktad till experter inom området. Tredje syftet uppnås genom en kvantitativ studie där formulerade hypoteser testas i en statistisk undersökning på 152 svenska börsnoterade bolag.  Resultat och slutsatser: Studien säkerställer ett approximativt mått för centraliseringsgraden av finansiering. Måttets samband med företagsvärde kan dock inte säkerställas. Centraliseringsgraden har däremot ett samband med företagsvärde under vissa förhållanden vilket innebär att ett samband inte heller kan uteslutas. Vidare finner studien empiriskt stöd för Tradeoff Theory, däremot kan ingen förklarande modell för kapitalstrukturen säkerställas och därmed inte heller centraliseringsgradens betydelse för kapitalstrukturen. Slutligen har studien även påbörjat en förklarande modell till centraliseringsgraden av finansiering i och med att geografisk spridning visar sig ha positiv påverkan på centraliseringsgraden. / Background: Firms using centralized financing function usually motivates it by it’s value adding property, something today’s research only theoretically touched which indicates that centralized financing can be a value adding strategy. Therefore it exists an interest for an empirical examination. The lack of a measurement for the degree of centralized financing implies that this measurement have to carried out before the theoretical value adding strategy can be empirical tested. Purpose: The study’s purposes is 1) bring forward a measurement for the degree of centralized financing, 2) secure the measurement’s validity and 3) through this measurement test it’s value adding effect to firm value. Method: The first purpose is achieved through an explorative research design where theories close to the subject together with observations of reality were applied. The second purpose is achieved through a smaller statistical study based on financial reports in conjunction with a survey aimed towards experts within the knowledge area. The third purpose is achieved through testing hypotheses in a statistical study on 152 Swedish listed companies. Results and conclusions: The study validates an approximate measure of the degree of centralized financing. The measurement’s connection to firm value can not be established. Although, the degree of centralization has a connection to firm value given certain circumstances which implies that the connection cannot be ruled out either. Further, the study finds empirical support for Tradeoff Theory, however, there are no empirical support for the explanatory model for the capital structure and, by extension, neither for the degree of centralization’s significance on capital structure. Finally, the study has also begun an explanatory model for the degree of centralization of financing since geographical spread has a significant positive effect on the degree of centralization.

In this thesis, we investigate several radio resourcemanagement (RRM) techniques and concepts in an indoorenvironment with a dense infrastructure. Future wireless indoorcommunication networks will very likely be implemented atplaces where the user concentration is very high. At these hotspots, the radio resources must be used efficiently. The goalis to identify efficient RRM techniques and concepts that aresuitable for implementation in an indoor environment. Handling the high level of co-channel interference is shownto be of paramount importance. Several investigations in thethesis point this out to be the key problem in an indoorenvironment with a dense infrastructure. We show that a locallycentralized radio resource management concept, the bunchconcept, can give a very high performance compared to othercommonly used RRM concepts. Comparisons are made withdistributed systems and systems using channel selection schemeslike CSMA/CA. The comparisons are primarily made by capacityand throughput analysis which are made by system levelsimulations. Results show that the centralized concept can give85 percent higher capacity and 70 percent higher throughputthan any of the compared systems. We investigate several RRM techniques to deal with thechannel interference problem and show that beamforming cangreatly reduce the interference and improve the systemperformance. Beamforming, especially sector antennas, alsoreduce the transmitter powers and the necessary dynamic range.A comparison is made between the use of TD/CDMA and pure TDMAwhich clearly shows the performance benefits of usingorthogonal channels that separates the users and reduces theco-channel interference. Different channel selection strategiesare studied and evaluated along with various methods to improvethe capability of system co-existence. We also investigate several practical measures to facilitatesystem implementation. Centralized RRM is suitable forguaranteeing QoS but is often considered too complex. With thestudied centralized concept the computational complexity can bereduced by splitting the coverage area into smaller pieces andcover them with one centralized system each. This reduces thecomplexity at the prize of lost capacity due to theuncontrolled interference that the different systems produce.Our investigations show that sector antennas can be used toregain this capacity loss while maintaining high reduction incomplexity. Without capacity loss, the computational complexitycan be reduced by a factor of 40 with sectoring. Theimplementation aspects also include installation sensitivity ofthe indoor architecture and the effect of measurement errors inthe link gains. The robustness against installation errors ishigh but the bunch concept is quite sensitive to largemeasurement errors in the studied indoor environment. Thiseffect can be reduced by additional SIR-margins of the radiolinks. The studied bunch concept is shown to be promising for usein future wireless indoor communication systems. It provideshigh performance and is feasible to implement. Keywords:Radio resource management, indoorcommunication, the bunch concept, centralized RRM, dynamicchannel allocation, channel selection, co-channel interference,power control, feasibility check, capacity, throughput, qualityof service, beamforming, downtilting, sector antennas,co-existence, computational complexity, sensitivity analysis,measurement errors, infrastructure, system implementation,WLAN, HiperLAN/2, IEEE 802.11.

Approved for public release, distribution is unlimited / MBA Professional Report / The purpose of project is to examine two alternatives among the numerous recommendations to improve United Nations Peacekeeping. The first calls for improved centralization of United Nations Peacekeeping. The second calls for unification of Peacekeeping operations through joint efforts (political, economical, cultural, religious, military, etc.) at the regional level in order to include the parties involved in the conflict and their neighboring states in Africa, America, Asia and Europe. The goal of this project is to identify and understand the debate about the changing role of the United Nations Peacekeeping and to identify ways to more effectively manage operations. / Lieutenant Colonel, Bulgarian Army

A demanda de largura de banda exigida pelos usuários de redes de acesso vem aumentando rapidamente e a rede óptica passiva baseada em multiplexação por divisão de comprimento de onda (WDM-PON) tem se destacado como a tecnologia capaz de suprir essa demanda. Dessa forma, este trabalho conduz, inicialmente, uma comparação entre uma rede WDM-PON ideal e uma rede TDM-PON, discutindo também aspectos de segurança desta última, bem como estratégias de migração entre estes dois esquemas. Devido ao custo elevado da WDM-PON em seu esquema original, investigou-se em seguida as principais arquiteturas alternativas relatadas na literatura, em busca de maior viabilidade de implementação. Com base nesse estudo, fez-se então uma análise mais profunda sobre topologias recentes derivadas da RITENET, referenciadas na literatura como arquiteturas baseadas em fontes ópticas centralizadas (CLS). Tais configurações são baseadas no travamento de diodos lasers Fabry-Pérot através da injeção do sinal de ASE de um EDFA e foram estudadas por meio de simulações computacionais empregando o software OptiSystem 8.0. / The demand for bandwidth by access networks users has increased rapidly and the Wavelength Division Multiplexing - Passive Optical Network (WDM-PON) has been pointed out as the technology capable of meeting this demand. In this framework, this work initially carries out a comparison between an ideal WDM-PON network and a TDM-PON network, also discussing TDM-PON security issues as well as migration strategies from TDM-PON to WDM-PON configurations. Next, taking into account the high implementation cost of the WDM-PON original scheme, we investigated alternative topologies available in the literature. Based on this study, we performed a more in-depth analysis regarding recently proposed topologies, derived from the RITENET, which are refered at the literature as centralized light sources (CLS) based architectures. Such networks are based on the locking of Fabry-Pérot lasers diodes through the injection of the ASE signal from an EDFA and were studied by using the software OptiSystem 8.0.

Orientador: Antonio Cesar Germano Martins / Resumo: Nos últimos anos a energia solar fotovoltaica tem mostrado forte crescimento dentre as fontes renováveis em várias partes do mundo. O Brasil vem discutindo mais intensamente desde 2012 formas de incentivo para a inserção da energia solar fotovoltaica bem como a adequação do modelo centralizado e distribuído na matriz elétrica nacional. Para contribuir com esta discussão, neste trabalho foi realizado um levantamanto de informações a respeito destes dois modelos a partir de consultas à livros, artigos, periódicos, sites e reportagens. Os resultados dessas pesquisas foram agrupados e consolidados em gráficos, tabelas e quadros que permitiram a avaliação das perspectivas da energia solar fotovoltaica no Brasil, identificando os principais aspectos dos dois modelos de geração, levando-se em consideração suas características distintas e relacionadas à geração, distribuição e consumo salientando-se as vantagens e desvantagens segundo critérios ambientais, econômicos; sociais e técnicos. As análises dos materiais constituintes da pesquisa denotam que, no caso do Brasil, os dois modelos progridem de maneira complementar, porém necessitam de políticas governamentais mais eficazes para as suas implementações de forma ampla e significativa na matriz energética nacional. / Mestre

The U.S. health care system is considered costly and highly inefficient, devoting substantial resources to the treatment of acute conditions in a hospital setting rather than focusing on prevention and keeping patients out of the hospital. The potential for cost savings is large; in the U.S. more than $30 billion are spent each year on hospitalizations deemed preventable, 31% of which is attributed to heart diseases and 20% to diabetes. Motivated by this, our work focuses on developing centralized and distributed learning methods to predict future heart- or diabetes- related hospitalizations based on patient Electronic Health Records (EHRs). We explore a variety of supervised classification methods and we present a novel likelihood ratio based method (K-LRT) that predicts hospitalizations and offers interpretability by identifying the K most significant features that lead to a positive prediction for each patient. Next, assuming that the positive class consists of multiple clusters (hospitalized patients due to different reasons), while the negative class is drawn from a single cluster (non-hospitalized patients healthy in every aspect), we present an alternating optimization approach, which jointly discovers the clusters in the positive class and optimizes the classifiers that separate each positive cluster from the negative samples. We establish the convergence of the method and characterize its VC dimension. Last, we develop a decentralized cluster Primal-Dual Splitting (cPDS) method for large-scale problems, that is computationally efficient and privacy-aware. Such a distributed learning scheme is relevant for multi-institutional collaborations or peer-to-peer applications, allowing the agents to collaborate, while keeping every participant's data private. cPDS is proved to have an improved convergence rate compared to existing centralized and decentralized methods. We test all methods on real EHR data from the Boston Medical Center and compare results in terms of prediction accuracy and interpretability.

O controle centralizado do tráfego tem como objetivo integrar e gerenciar informações, auxiliando na tomada de decisão em tempo real. Devido ao aumento da complexidade da malha viária, principalmente em grandes cidades, muitos estudos têm focado em sistemas de controle de tráfego, incluindo desenvolvimento de novas tecnologias e ferramentas. Para gerenciar essa complexidade, é necessário que as organizações públicas, onde estão inseridos os Centros de Controle de Tráfego (CCT), conheçam e melhorem seus processos, alinhando e integrando-os aos seus sistemas de informação, de maneira a possibilitar o atendimento das demandas de forma eficiente. Nesse sentido, esta pesquisa baseia-se nos conceitos e práticas do Business Process Management (BPM) com objetivo de propor um modelo de estruturação dos processos para a gestão inteligente das informações no controle centralizado do tráfego. O trabalho, portanto, se desenvolve ao longo das fases e etapas do ciclo do BPM, contemplando desde o planejamento até a modelagem de processos e apresentando como resultado final uma proposta para um plano de otimização de processos. Assim, para atender ao objetivo principal da pesquisa, o trabalho foi desmembrado em três artigos com níveis crescentes de glanuralidade: (i) Inicialmente foi apresentada, baseada em estudo de caso, uma aplicação abrangente da metodologia, percorrendo da primeira à terceira fase do ciclo de BPM em que a empresa está apta para executar os processos e seguir para a quarta fase, Controle e Análise de Dados. (ii) O segundo artigo apresenta o detalhamento da segunda fase do ciclo de BPM, que consiste na modelagem e otimização dos processos. Os processos As Is foram, então, modelados e analisados para que pudesse ser proposta a melhoria na forma de um macroprocesso To Be. (iii) Para embasar a proposição do macroprocesso To Be, foi realizada uma revisão sistemática de literatura, na qual foram compiladas as melhores práticas no tema e verificadas para a realidade brasileira através de entrevista com especialistas. O resultado foi o Mapeamento de Melhores Práticas associado a um Modelo Referência para o Processo de Controle de Tráfego. Em complemento, foram determinadas diretrizes a serem seguidas para a gestão inteligente das informações no controle centralizado do tráfego. Esses resultados definem, portanto, um modelo de referência To Be dos processos de trabalho em CCTs. A principal contribuição teórica desta pesquisa é, então, a consolidação de melhores práticas associadas a um Modelo de Referência do Processo de Controle do Tráfego, ajustado para a realidade dos CCTs brasileiros. Do ponto de vista prático, a metodologia e os resultados apresentados incentivam a implantação de BPM não só em CCTs, mas em qualquer setor de uma organização pública. / The centralization of traffic control aims to integrate and manage information, aiding decision making in real time. Due to the increasing complexity of the road network, especially in large cities, many studies have focused on traffic control systems, including development of new technologies and tools. To manage this complexity, it is necessary that public organizations, where the Traffic Control Centers (TCC) are inserted, understand and improve their processes, aligning and integrating them into their information systems in order to enable the fulfillment of demands efficiently. In this sense, this research is based on the concepts and practices of Business Process Management (BPM) in order to propose a structuring model of processes for the intelligent management of information in centralized traffic control. The work, therefore, is developed through the phases and stages of the BPM cycle, covering from planning to process modeling and presenting as the final result a proposal for a process optimization plan. Thus, to meet the main goal of the research, the work was divided into three articles with increasing levels of detail: (i) initially it was presented, based on case study, a comprehensive application of the methodology, that cover from the first to the third phase of the BPM cycle in which the company is able to run the process and move on to the fourth phase, Control and Data Analysis. (ii) The second article presents the details of the application of second phase of the BPM cycle, which consists of modeling and optimization of processes. The processes "As Is" were modeled and analyzed in order to propose improvements trough a Macro process "To Be". (iii) To support the proposition of macroprocess “To Be”, a systematic review of the literature was applied in order to compile best practices on the subject that were checked for the Brazilian reality through interviews with experts. The result was a Map of Best Practices in association with a Reference Model for Traffic Control Process. In addition to the map, the paper presents guidelines to be followed for the intelligent management of information in centralized traffic control. These results define a reference model "To Be" of work processes in TCCs. From a practical standpoint, the methodology and results presented encourage the deployment of BPM not only in TCCs, but also in any department of a public organization.

The paper aims to investigate the effects of ¡§Local Government Act¡¨on the ¡§centralized personnel system¡¨ of the government employee ethics unit. Before the promulgation and implementation of the Local Government Act, the laws of local self-government for the local government agencies to follow and observe were ¡§The Taiwan Local Self-Government Guidelines¡¨, ¡§The Self-Government Law for Provinces and Counties¡¨, and ¡§The Municipal Self-Government Law¡¨. In fact, the local self-government implemented by the county and municipal governments only existed in name because the so-called local self-government was still under the control of the Central Government. Before the government employee ethics unit was reorganized in 1992, the government employee ethics officers were put in the government departments of different levels as well as the state-run enterprises in Taiwan, and subordinated to the personnel units of these institutions. They were responsible for maintaining the safety and security of the institutions, investigating the loyalty of the staff members, and so on. After the implementation of Local Government Act, the chiefs of local government agencies elected by the people have been vested much more personnel authorities. They also expect a higher level of autonomy as the self-consciousness of the people is growing high. Along with this reorganization, the government employee ethics officers have turned into the staff members to assist the chiefs of local government agencies. Meanwhile, based on their statutory job duties, these staff members concurrently assume the great responsibilities for well-formed civil service ethics and anti-corruption practices for local government agencies.

Background: Rottne Industri is a Swedish heavy machinery manufacturer whose production process is divided in three manufacturing plants. The production takes place in the facilities of Stensele and Lenhovda while the one at Rottne is also used for the final assembly. Ten suppliers provide steel materials to each facility translating to relatively high transportation costs which may be reduced by adopting a more centralized structure. Purpose: The purpose of this paper is to identify the transportation costs between the steel suppliers and the three manufacturing plants at Stensele, Lenhovda and Rottne as well as the freight transportation costs for components from Stensele and Lenhovda to Rottne. Moreover, the research seeks to analyze the difference in transportation costs if Rottne Industri was to centralize all its activities into one single manufacturing facility at Rottne. Finally, the authors aim to identify the environmental cost related to freight transportation in order to evaluate the total transportation cost difference between the two scenarios. Method: The Master thesis focuses on an instrumental case study on a single company. Qualitative and quantitative research methods are used for the necessary data collection to bring up a broader picture of Rottne Industri transportation costs and how it will change in a different organizational structure. Results: The studied company would reduce its freight transportation cost from 2 471 735 SEK of the current decentralized structure to 398 265 SEK of the centralized one. The environmental cost would decrease by 91% (20 420 SEK) due to the decrease of material flow. Therefore Rottne Industri would reduce its total transportation costs by 84% by centralizing its manufacturing structure, for a monetary value estimated at 2 095 860 SEK. However these results contradict part of the theoretical framework concerning the advantages and the disadvantages of centralization and decentralization, as the new centralized structure appears to be more easily manageable while reducing the transportation costs from the different suppliers. Therefore further researches are required as a broader approach on the effects of switch in the organizational structure at Rottne Industri.

This thesis describes the work performed at Saab Aeronautics in Linkoping during 2011. The work was to study and develop distributed and centralized methods for analysis of the decision making in a group of unmanned aerial vehicles (UAVs). In addition to this, some simple scenarios were studied and the work was implemented in C++ as the simulator ComDec. The literature review presented in the report provides knowledge of strategies for collaborative UAVs (communication and decision making), understanding of the problems/constraints that are relevant for data links and insight into the algorithms for decision making and autonomy. This work has resulted in a theoretical analysis of suitable design for decision making in a group of interacting autonomous UAVs. Existing methods for distributed and centralized decision making are implemented and a demonstration ofthe outcome are presented in the thesis. It is further shown how various communication problems and disturbances affect the decision making process. Finally, advantages and disadvantages of the selected strategies (communication and decision making) are discussed.