Submitted by Oliver Sleczka (oliver.sleczka@student.unisg.ch) on 2017-12-11T13:34:22Z No. of bitstreams: 1 Performance of Private Equity Funds in Emerging Markets_MPGI Dissertation_2017.pdf: 1511526 bytes, checksum: 5880b1437aa5a15cc2e08f09ac374e4a (MD5) / Rejected by Josineide da Silva Santos Locatelli (josineide.locatelli@fgv.br), reason: Dear Oliver, There are some corrections to do on your thesis, please, see below: The title is different: PERFORMANCE OF PRIVATE EQUITY FUNDS IN EMERGING MARKETS An Empirical Analysis We have: PERFORMANCE OF PRIVATE EQUITY FUNDS IN EMERGING MARKETS - A COMPARATIVE ANALYSIS Page 2: In “Knowledge Field”, must to be your advisor field: Economia e Finanças Internacionais; Page 4: In “Knowledge Field”, must to be your advisor field: Economia e Finanças Internacionais, in “Approval Date” put your presentation date; Page 5: “ACKNOWLEDGMENT” must to be on the middle of the page; After correction, please, post it again on 2017-12-11T14:24:20Z (GMT) / Submitted by Oliver Sleczka (oliver.sleczka@student.unisg.ch) on 2017-12-11T17:40:14Z No. of bitstreams: 1 Performance of Private Equity Funds in Emerging Markets_MPGI Dissertation_2017.pdf: 1511283 bytes, checksum: 77cd6246c673012b4ee5c97c532f3274 (MD5) / Approved for entry into archive by Josineide da Silva Santos Locatelli (josineide.locatelli@fgv.br) on 2017-12-11T17:46:14Z (GMT) No. of bitstreams: 1 Performance of Private Equity Funds in Emerging Markets_MPGI Dissertation_2017.pdf: 1511283 bytes, checksum: 77cd6246c673012b4ee5c97c532f3274 (MD5) / Made available in DSpace on 2017-12-11T17:50:20Z (GMT). No. of bitstreams: 1 Performance of Private Equity Funds in Emerging Markets_MPGI Dissertation_2017.pdf: 1511283 bytes, checksum: 77cd6246c673012b4ee5c97c532f3274 (MD5) Previous issue date: 2017 / This paper investigates the performance of private equity funds in emerging markets based on data of individual fund level returns from 1995 to 2013 obtained from Preqin. The research focus of this thesis lies on performance drivers and return persistence. The analysis is divided in two parts. The first part investigates the relationship between fund performance and fund characteristics focusing on the GP location, the investment focus, as well as the fund size, sequence number, and GP experience. The key objective is to determine significant emerging market-specific performance factors. Our results suggest that funds with a clear geographic investment focus earn significantly higher returns. The second part aims to find evidence for persistence in the performance of emerging market private equity funds. The findings corroborate the hypothesis that unlike in developed markets, performance persistence still exists in emerging markets. / Esta tese investiga o desempenho de fundos de private equity em mercados emergentes, com base em dados de retorno a nível individual dos fundos de 1995 a 2013, obtidos por meio da base de dados Preqin. O foco de pesquisa desta tese é nos drivers de performance e persistência de retornos. A análise é dividida em duas partes. A primeira parte investiga a relação entre a performance e características do fundo, com foco na localização do administrador (GP), foco do investimento, assim como o tamanho do fundo, número de sequência do fundo, e experiência do GP. O principal objetivo é determinar fatores de performance significativos específicos de mercados emergentes. Os resultados sugerem que fundos com um claro foco geográfico de investimentos obtém retornos significativamente mais altos. A segunda parte visa encontrar evidências de persistência na performance de fundos de private equity de mercados emergentes. Os achados corroboram a hipótese de que, diferentemente de mercados desenvolvidos, persistência de performance ainda existe em mercados emergentes.

Les grandes évolutions de la technique de systèmes embarqués au cours des dernières années ont permis avec succès la combinaison de la détection, le traitement des données, et diverses technologies de communication sans fil tout en un nœud. Les réseaux de capteurs sans fil (WSN) qui se composent d’un grand nombre de ces nœuds ont attiré l’attention du monde entier sur les établissements scolaires et les communautés industrielles, puisque leurs applications sont très répandues dans des domaines tels que la surveillance de l’environnement, le domaine militaire, le suivi des événements et la détection des catastrophes. En raison de la dépendance sur la batterie, la consommation d’énergie des réseaux de capteurs a toujours été la préoccupation la plus importante. Dans cet article, une méthode mixte est utilisée pour l’évaluation précise de l’énergie sur les réseaux de capteurs, ce qui inclut la conception d’un environnement de SystemC simulation base au niveau du système et au niveau des transactions pour l’exploration de l’énergie, et la construction d’une plate-forme de mesure d’énergie pour les mesures de nœud banc d’essai dans le monde réel pour calibrer et valider à la fois le modèle de simulation énergétique de nœud et le modèle de fonctionnement. La consommation d’énergie élaborée de plusieurs différents réseaux basés sur la plate-forme de nœud sont étudiées et comparées dans différents types de scénarios, et puis des stratégies globales d’économie d’énergie sont également données après chaque scénario pour les développeurs et les chercheurs qui se concentrent sur la conception des réseaux de capteurs efficacité énergétique. Un cadre de l’optimisation basée sur un algorithme génétique est conçu et mis en œuvre à l’aide de MATLAB pour les réseaux de capteurs conscients de l’énergie. En raison de la propriété de recherche global des algorithmes génétiques, le cadre de l’optimisation peut automatiquement et intelligemment régler des centaines de solutions possibles pour trouver le compromis le plus approprié entre la consommation d’énergie et d’autres indicateurs de performance. Haute efficacité et la fiabilité du cadre de la recherche des solutions de compromis entre l’énergie de nœud, la perte de paquets réseau et la latence ont été prouvés par réglage paramètres de l’algorithme CSMA / CA de unslotted (le mode non-beacon de IEEE 802.15.4) dans notre simulation basé sur SystemC via une fonction de coût de la somme pondérée. En outre, le cadre est également disponible pour la tâche d’optimisation basée sur multi-scénarios et multi-objectif par l’étude d’une application médicale typique sur le corps humain. / The great technique developments of embedded system in recent years have successfully enabled the combination of sensing, data processing and various wireless communication technologies all in one node. Wireless sensor networks (WSNs) that consist of many of such node have gained worldwide attention from academic institutions and industrial communities, since their applications are widespread in such as environment monitoring, military fields, event tracing/tracking and disaster detection. Due to the reliance on battery, energy consumption of WSNs has always been the most significant concern. In this paper, a mixed method is employed for the accurate energy evaluation on WSNs, which involves the design of a transaction-level system-level based SystemC simulation environment for energy exploration, and the building of an energy measurement system platform for the real world testbed node measurements to calibrate and validate both node energy simulation model and operation model. Elaborate energy consumption of several different node platform based networks are investigated and compared under different kinds of scenarios, and then comprehensive energy-saving strategies are also given after each case scenario for the developers and researchers who focus on the energy-efficient WSNs design. A genetic algorithm (GA) based optimization framework is designed and implemented using MATLAB for the energy aware WSNs. Due to the global search property of genetic algorithms, the optimization framework is able to automatically and intelligently fine tune hundreds/thousands of potential solutions to find the most suitable tradeoff among energy consumption and other performance metrics. The framework’s high efficiency and reliability of finding the tradeoff solutions among node energy, network packet loss and latency have been proved by tuning unslotted CSMA/CA algorithm parameters (used by non-beacon mode of IEEE 802.15.4) in our SystemC-based simulation via a weighted sum cost function. Furthermore, the framework is also available for the multi-scenario and multi-objective based optimization task by studying a typical medical application on human body. Keywords: Wireless sensor networks (WSNs), energy consumption, simulation/emulation, SystemC, testbeds, measurements, calibration, optimization, genetic algorithms, performance metrics, weighted sum cost function, multi-scenario and multi-objective optimization, Pareto-front

No description available.

[pt] O sucesso operacional de uma empresa depende muito da maneira como a cadeia de suprimentos é gerenciada. Em razão do seu papel estratégico, a gestão adequada pode proporcionar uma série de benefícios, aumentando sua competitividade e computando a rentabilidade de seus negócios. Esta dissertação avalia por meio de um estudo de caso, a cadeia de suprimentos de bens da Petrobras, especificamente na área de relacionamento com fornecedores e estratégias de contratações de bens. Os processos de suprimentos, especificamente na área de compras de bens, foram analisados utilizando como base a metodologia SCOR em relação a processos de compras, prazos, custos de materiais, canais de suprimentos e planejamento estratégico e então foram identificados os principais gargalos (gaps) e riscos de processos relacionados a padrões internos, exigências legais e relacionamento com o mercador fornecedor. Também foram indicadas melhorias futuras que podem ser implementadas nestes processos de forma a contribuir para melhorias de eficiência em processos de compras da cadeia de suprimentos de bens e relacionamento com o mercado. Além da contribuição acadêmica, este trabalho permite esclarecer ao meio empresarial sobre os benefícios que a metodologia SCOR pode trazer para o gerenciamento de cadeias de suprimentos, além das práticas de aperfeiçoamento operacional. Por fim, neste estudo de caso, foram identificadas oportunidades de melhorias baseadas nos gargalos de processos identificados como Current Disconnects e definidas as melhorias futuras (Future state opportunities) para os processos de relacionamentos com fornecedores e estratégias de contratações. / [en] A company s operational success depends a lot on how the supply chain is managed. Due to its strategic role, proper management can provide a series of benefits, increasing your competitiveness and computing the profitability of your business. This dissertation evaluates through a case study, the supply chain of Petrobras goods, specifically in the area of relationship with suppliers and strategies for contracting goods. The supply processes, specifically in the area of purchasing goods, were analyzed using the SCOR methodology as a basis in relation to purchasing processes, deadlines, material costs, supply channels and strategic planning, and then the main bottlenecks (gaps) were identified. and process risks related to internal standards, legal requirements and relationship with the supplier merchant. Future improvements were also indicated that can be implemented in these processes in order to contribute to efficiency improvements in purchasing processes in the goods supply chain and relationship with the market. In addition to the academic contribution, this work provides clarification to the business community about the benefits that the SCOR methodology can bring to the management of supply chains, in addition to operational improvement practices. Finally, in this case study, improvement opportunities were identified based on process bottlenecks identified as Current Disconnects, and future state opportunities were defined for supplier relationship processes and hiring strategies.

Radar is an invaluable tool for detecting and assessing threats on land, on the seas and in the air. To properly evaluate threats, radar operators construct threat libraries where the signal characteristics of emitters are stored and mapped to specific types of platforms. In this project, methods for constructing these threat detection libraries from data obtained during real-life scenarios are investigated. A number of machine learning approaches are investigated and validated using general and method specific scoring methods. Using density based clustering methods and non-linear data transformation it is shown that Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) and spatial consistency metrics can be used to deinterleave and group signals to radar trace emitting platforms, from which suitable library parameters can be extracted. The results show that traditional metrics for evaluating cluster methods are not suited for evaluating data containing spatial information. / Radar är ett ovärderligt verktyg för att upptäcka och identifiera hot på land, till havs och i luften. För att kunna utvärdera olika former av hot använder sig radaroperatörer av hotbibliotek, vilka består av olika radarplattformers signalparametrar. I det här projektet undersöks olika metoder för att bygga hotbibliotek med hjälp av verkliga data insamlat under flygningar i Sverige. Olika maskininlärningsmetoder undersöks och utvärderas med hjälp av både generella och specifika utvärderingsmetoder. Genom att använda sig av densitets- baserade klustringsmetoder och olinjära metoder för att transformera data så visas att hierarkisk densitetsbaserad spatial klustring för tillämningar med störningar (HDBSCAN) och utvärderingsmetoder som baseras på spatial karaktäristik kan användas för att separera och gruppera radarkällor, vilka kan användas för att finna parametrar för att bygga hotbibliotek. Det visas även att traditionella metoder för att utvärdera klustringsresultat inte lämpar sig för att utvärdera spatiala data.

Цель работы – разработка, реализация и внедрение системы управления IT-услугами, позволяющей повысить качество и скорость работы технического отдела компании. Объект исследования – класс ITSM-систем, позволяющих управлять работой IT-отдела компании. Методы исследования: анализ и сравнение существующих систем управления IT-услугами; анализ современных средств разработки WEB-приложений; тестирование разрабатываемой системы на реальных пользователях и реальных данных. Результаты работы: разработана система управления IT-услугами (ITSM) для компании по предотвращению магазинных краж. Доказана эффективность внедрения новой системы. Выпускная квалификационная работа выполнена в текстовом редакторе Microsoft Word и представлена в твёрдой копии. / The goal of the work is to develop, implement and implement an IT service management system that will improve the quality and speed of work of the company’s technical department. The object of study is a class of ITSM systems that allow you to manage the work of a company’s IT department. Research methods: analysis and comparison of existing IT service management systems; analysis of modern WEB application development tools; testing the developed system on real users and real data. Work results: an IT service management (ITSM) system was developed for a shoplifting prevention company. The effectiveness of implementing the new system has been proven. The final qualifying work was completed in the text editor Microsoft Word and presented in a hard copy.

Performance measurement and runtime tuning tools are both vital in the HPC software ecosystem and use similar techniques: the analyzed application is interrupted at specific events and information on the current system state is gathered to be either recorded or used for tuning. One of the established performance measurement tools is Score-P. It supports numerous HPC platforms and parallel programming paradigms. To extend Score-P with support for different back-ends, create a common framework for measurement and tuning of HPC applications, and to enable the re-use of common software components such as implemented instrumentation techniques, this paper makes the following contributions: (I) We describe the Score-P metric plugin interface, which enables programmers to augment the event stream with metric data from supplementary data sources that are otherwise not accessible for Score-P. (II) We introduce the flexible Score-P substrate plugin interface that can be used for custom processing of the event stream according to the specific requirements of either measurement, analysis, or runtime tuning tasks. (III) We provide examples for both interfaces that extend Score-P’s functionality for monitoring and tuning purposes.

This thesis aims to optimize the machine learning algorithms for predicting KPI metrics for an organization. The organization is predicting whether projects meet planned deadlines of the last phase of development process using machine learning. The work focuses on the analysis of prediction models and sets the goal of selecting new candidate models for the prediction system. We have implemented a system that automatically selects the best feature variables for learning. Trained models were evaluated by several performance metrics and the best candidates were chosen for the prediction. Candidate models achieved higher accuracy, which means, that the prediction system provides more reliable responses. We suggested other improvements that could increase the accuracy of the forecast.

Performance measurement and runtime tuning tools are both vital in the HPC software ecosystem and use similar techniques: the analyzed application is interrupted at specific events and information on the current system state is gathered to be either recorded or used for tuning. One of the established performance measurement tools is Score-P. It supports numerous HPC platforms and parallel programming paradigms. To extend Score-P with support for different back-ends, create a common framework for measurement and tuning of HPC applications, and to enable the re-use of common software components such as implemented instrumentation techniques, this paper makes the following contributions: (I) We describe the Score-P metric plugin interface, which enables programmers to augment the event stream with metric data from supplementary data sources that are otherwise not accessible for Score-P. (II) We introduce the flexible Score-P substrate plugin interface that can be used for custom processing of the event stream according to the specific requirements of either measurement, analysis, or runtime tuning tasks. (III) We provide examples for both interfaces that extend Score-P’s functionality for monitoring and tuning purposes.

[pt] Em muitos países, uma área de Exploração e Produção de petróleo é adquirida através de um leilão. Embora o processo de liquidação do leilão seja tipicamente simples, sob a ótica do tomador de decisão a identificação da melhor oferta é complexa. Para sua valoração, deve ser pré-definido o modelo de desenvolvimento com diversas alternativas associadas, alto grau de incertezas técnicas, de mercado e operacionais, e submetido a uma determinada condição fiscal. A decisão por uma determinada alternativa gera impactos e investimentos elevados para a empresa. O trabalho proposto visa a construção de um framework para dar suporte ao processo de escolha da melhor oferta que maximize uma medida de valor para a empresa, auxiliando o tomador de decisão e considerando as incertezas envolvidas no processo. Foram utilizados os indicadores de performance apresentados na literatura: Valor Presente Líquido (VPL), Conditional Value-at-Risk, Omega e Exposição Financeira. Afim de melhor quantificar risco/benefício financeiro no processo de tomada de decisão foram construídas medidas de risco: Mean-Weighted CVaR, Mean-Weighted Double-Sided CVaR, Beta e Negative-Positive cashflow ratio. Para ilustrar a aplicabilidade do framework proposto, um experimento numérico baseado em um caso hipotético é apresentado. Em decorrência deste experimento, foi identificado que alterações na configuração da produção alteram significativamente os resultados dos indicadores. Além disso, a partir de uma ponderação entre probabilidade de ganho e o resultado do indicador Mean-Weighted CVaR, foi identificada a melhor oferta para a área, dada a condição fiscal do leilão. / [en] In many countries, an oil and gas area of Exploration and Production is acquired through an auction. Although the auction settlement process is typically simple, from the decision maker s point of view, identifying the best offer is complex. For its valuation, the development model must be pre-defined with several associated alternatives, a high degree of technical, market and operational uncertainties, and submitted to a fiscal term. The decision for a particular alternative generates high impacts and investments for the company. The proposed work aims to build a framework to support the process of choosing the best offer that maximizes a measure of value for the company, helping the decision maker and considering the uncertainties involved in the process. The performance indicators presented in the literature were used: Net Present Value (NPV), Conditional Value-at-Risk, Omega and Financial Exposure. In order to better quantify financial risk/benefit in the decisionmaking process, risk measures were constructed: Mean-Weighted CVaR, Mean- Weighted Double-Sided CVaR, Beta and Negative-Positive cashflow ratio. To illustrate the applicability of the proposed framework, a numerical experiment based on a hypothetical case is presented. As a result of this experiment, it was identified that changes in the configuration of production significantly alter the results of the indicators. In addition, from a weighting between the probability of gain and the result of the Mean-Weighted CVaR measure, the best offer for the area was identified, given the fiscal term of the bid.