A new domain of commercial applications demands the development of inexpensive parallel computing platforms to lower the cost of operations and increase the business profit. The calculation of returns on an IT investment is now important to justify the decision of upgrading or replacing parallel systems. This thesis presents a framework of the performance and economic factors that are considered when evaluating a parallel system. We introduce a metric called the cost/profit effective metric, which measures the effectiveness of a parallel system in terms of performance, cost and profit. This metric describes the profit obtained from the performance of three different domains for scaling: speed-up, throughput and/or scale-up. Cost is measured by the actual costs of a parallel system. We present two cases of study to demonstrate the application of this metric and analyze the results to support the evaluation of the parallel system on each case.

No description available.

No description available.

No description available.

No description available.

No description available.

La prévision de la profitabilité du client, ainsi que la tarification, sont des pièces centrales dans le monde des sciences actuarielles. En utilisant des données sur les historiques des clients et en optimisant des modèles statistiques, les actuaires peuvent prévoir, dans une certaine mesure, le montant qu’un client réclamera durant une certaine période. Cependant, ces modèles utilisent souvent des données sensibles reliées au client qui sont considérées comme étant des facteurs de risque très importants dans la prédiction de pertes futures. Ceci est considéré comme étant légal dans plusieurs jurisdictions tant que leur utilisation est supportée par des données actuarielles, car ces attributs permettent aux clients d’obtenir une prime plus précise. Toutefois, comme soulevé dans la littérature récente en apprentissage machine, ces modèles peuvent cacher des biais qui les rendent discriminants envers certains groupes. Dans ce mémoire, nous proposons un modèle de prévision de la profitabilité du client utilisant des avancées récentes provenant du domaine de l’apprentissage machine pour assurer que ces algorithmes ne discriminent pas disproportionnellement envers certains sous-groupes faisant partie de l’intersection de plusieurs attributs protégés, tel que l’âge, la race, la religion et l’état civil. En d’autres mots, nous prédisons équitablement la prime théorique de n’importe quel client en combinant l’état de l’art en prédiction de pertes en assurance et appliquant certaines contraintes d’équité sur des modèles de régression. Suite à l’exécution de l’estimation de la profitabilité du client sur plusieurs jeux de données réels, les résultats obtenus de l’approche proposée sont plus précis que les modèles utilisés traditionnellement pour cette tâche, tout en satisfaisant des contraintes d’équité. Ceci montre que cette méthode est viable et peut être utilisée dans des scénarios concrets pour offrir des primes précises et équitables aux clients. Additionnellement, notre modèle, ainsi que notre application de contraintes d’équité, s’adapte facilement à l’utilisation d’un grand jeu de données qui contiennent plusieurs sous-groupes. Ceci peut être considérable dans le cas où un critère d’équité intersectionnel doit être respecté. Finalement, nous notons les différences entre l’équité actuarielle et les définitions d’équité provenant du monde de l’apprentissage machine, ainsi que les compromis reliés à ceux-ci. / Customer profitability forecasting, along with ratemaking, are central pieces in the world of actuarial science. By using historical data and by optimising statistical models, actuaries can predict whether a client with certain liabilities will claim any loss and what amount will be claimed inside a defined policy period. However, these models often use sensitive attributesrelated to the customer that are considered to be crucial risk factors to consider in predicting future losses. This is considered legal in many jurisdictions, as long as their use is backedby actuarial data, as these attributes give a more accurate premium to clients. Nonetheless,as it has been noted in recent machine learning literature, models can hide biases that makethem discriminate against certain groups. In this thesis, we propose a customer profitability forecasting model that uses recent advancements in the domain of machine learning to ensurethat these algorithms do not discriminate disproportionately on a subgroup of any intersectionof protected attributes, such as age, gender, race, religion and marital status. In other words,we fairly predict the theoretical premium of any client by combining state-of-the-art methodsin insurance loss prediction and the application of fairness constraints on regression models. After performing customer profitability estimation on multiple real world datasets, it is shownthat the proposed approach outperforms traditional models usually used for this task, whilealso satisfying fairness constraints. This shows that this method is viable and can be used inreal world scenarios to offer fair and accurate premiums to clients. Additionally, our model andour application of fairness constraints scale easily when using large datasets that contain many subgroups. This can be substantial in the case of satisfying an intersectional fairness criterion.Finally, we highlight the differences between actuarial fairness and fairness definitions in theworld of machine learning, along with its related trade offs.

Doctor of Education / Department of Educational Leadership / W. Franklin Spikes / Non-profit organizations come in many shapes and sizes and in some cases, encompass a significant portion of community organizations. Since these agencies are largely sustained by the work of volunteers, it is a worthy subject to examine. The diversity and complexity of non-profits available among communities presents a need for volunteers’ to be skilled, trained, and experienced in an array of arenas. Non-profit organizations depend on volunteers and allow these entities to reach a larger population than exclusively with the limited staff members available (Fisher & Cole, 1993). This notion suggested the topic of motivation and retention of volunteers among community non-profits for the focus of this study to help organizations effectively deliver their mission. More specifically, this investigation centered on people who are considered high-functioning by the administrators with whom they work through the process of interviews. Non-profit organizations located in Denver, Colorado were selected to participate. The administrators of the chosen agencies were contacted and asked to identify volunteers in their program they classify as high-functioning. Once this list was identified, these individuals were contacted and asked to participate in an interview. This study focused on high-functioning volunteers for data collection because it was anticipated their characteristics and viewpoints will aid in the understanding of motivation and retention factors for anyone working with and supervising volunteers in the 21st century. The findings from the study provide a diverse assessment of the topic of volunteer motivation and retention. The data collected suggests that adult volunteers have varying needs and the importance for volunteer administrators to take these differences into consideration in their management and programming efforts.

Submitted by Adriana Alves Rodrigues (aalves@espm.br) on 2017-11-21T14:30:14Z No. of bitstreams: 1 Ivan_Rizzo.pdf: 2595347 bytes, checksum: 1586a46a8ae1843dc9e9692ee54c0dec (MD5) / Approved for entry into archive by Adriana Alves Rodrigues (aalves@espm.br) on 2017-11-21T14:30:36Z (GMT) No. of bitstreams: 1 Ivan_Rizzo.pdf: 2595347 bytes, checksum: 1586a46a8ae1843dc9e9692ee54c0dec (MD5) / Approved for entry into archive by Ana Cristina Ropero (ana@espm.br) on 2017-11-22T12:59:20Z (GMT) No. of bitstreams: 1 Ivan_Rizzo.pdf: 2595347 bytes, checksum: 1586a46a8ae1843dc9e9692ee54c0dec (MD5) / Made available in DSpace on 2017-11-22T12:59:35Z (GMT). No. of bitstreams: 1 Ivan_Rizzo.pdf: 2595347 bytes, checksum: 1586a46a8ae1843dc9e9692ee54c0dec (MD5) Previous issue date: 2017-03-13 / Entidades esportivas tais quais clubes, federações, ONGs (Organizações Não Governamentais) e outras formas de organizações esportivas não lucrativas se tornaram organizações de interesse público mundial e não podem mais ser consideradas como fenômenos meramente locais, sendo reconhecidas tanto em mercados tradicionais (como Europa) ou mercados emergentes como Brasil, China e Índia e novas praças para o futebol como a América do Norte. Entre tanto, a grande disparidade entre clubes que pertencem aos principais mercados do futebol (chamados de “Big Five – Alemanha, Espanha, França, Inglaterra e Itália) representa um potencial de distanciar-se de equipes de outras ligas que pode afetar a capacidade de estes outros times investir além de seus mercados originais. O objetivo esta pesquisa é identificar os principais fatores para a internacionalização de uma equipe de futebol brasileira, por meio da abordagem teórica de redes estratégicas e visão institucional e identificar a dinâmica entre estes fatores, por meio da proposição de um modelo conceitual que descreve o desempenho de Organizações Esportivas Não Lucrativas. A revisão da literatura revela que os laços e conexões que permitem acesso a recursos inimitáveis, corridas de aprendizado e tornarse líder ou estar em posição estratégica em determinada rede estratégica são os principais fatores para a formação de uma rede visando a internacionalização de determinada entidade esportiva. Para atingir este objetivo, esta pesquisa foi realizada por meio de uma pesquisa exploratória qualitativa em estudo de caso único, utilizando a organização Sport Club Corinthians Paulista como objeto, com dados coletados via entrevistas semiestruturadas com 15 (quinze) atores-chave da organização (entre gestores e ex-gestores, jornalistas e parceiros de negócio). Os principais achados da pesquisa e o modelo proposto ao final desta pesquisa consideram a capacidade de aquisição de recursos de terceiros e as capacidades organizacionais internas como vetores do desempenho superior, além do tamanho da organização, sua imagem ou reputação além de fatores institucionais são fatores relevantes para a internacionalização de uma organização esportiva em redes estratégicas. Como contribuição teórica, organizações com marcas ou imagem superiores à média são fatores relevantes para a criação ou entrada em uma rede estratégica. Em relação à contribuição gerencial, destaca-se uma sequência de fatores que viabilizam a internacionalização de uma organização esportiva que visa audiência global. / Sports entities such as clubs, federations, NGOs (Non-Governmental Organizations), and other forms of nonprofit sports organizations (NPSOs) can no longer be considered as local phenomenon to become organizations known broadly all over the world and public interest and awareness can be perceived not only in Brazilian market, as in traditional markets (such as European audience) and on emerging markets as China and India, new markets to football such as North America. However, the great disparity among clubs that belong to the called “Big Five” (English, French, German, Italian and Spanish Leagues) has potential to develop such a great distance to teams from other leagues that can have an effect on these teams capacity to invest beyond their original markets. The objective of this research is to identify the main factors for Brazilian football team internationalization by strategic networks and institutional view approaches by proposing a conceptual model which describes NPSO’s performance. Literature review reveals that ties and connections to inimitable resources, learning races and being a leader or to be in an important position in network are main factors for network formation in order to achieve internationalization. For that purpose it was adopted a qualitative exploratory research in single case study, with a Brazilian NPSO, Sport Club Corinthians Paulista – in this research called Corinthians, data was collected by in depth-interviews with fifteen key respondents. These depth-interviews were conducted with Corinthians’ management, business partners, as well as sports management journalists from major Brazilian media companies. At the end of this research it is proposed a performance model for nonprofit sports organizations, which considers the resource acquisition capacity, internal capabilities as antecedents to its performance. Main findings revealed that resource acquisition capabilities, organization`s size, brand and image reputation, in addition of institutional environmental are relevant factors for internationalization by networks. As a theoretical contribution, firms having strong brand and image reputation with resources that will strength network are an important factor for a firm entering a network or to form one. Regarding managerial perspective, a main contribution is to propose a sequence of factors for sports organizations aiming to achieve global audience.

O presente trabalho tem como objetivo estudar a contabilidade no segmento de Incorporação e Construção Civil, como geradora de informações para a avaliação de empreendimentos no Longo Prazo, comparando o método contábil e o econômico. O estudo faz o levantamento do cenário econômico, inserindo o segmento em questão, apresentando as regras de negócios. Na etapa posterior busca-se de forma breve, traçar um histórico da contabilidade no mundo e no Brasil, inserindo-a no segmento, com a aderência as normas internacionais e seus impactos. São apresentadas as análises dos demonstrativos contábeis com a escolha de determinados indicadores para o estudo de caso, na seção posterior são apresentados os métodos de análise de investimentos, abordando o custo de oportunidade e duas opções para mensuração desse custo, evidenciando as diferenças entre os lucros contábil e econômico. Na penúltima seção há a apresentação de um estudo de caso para comparação dos resultados. / The present thesis studies the Accounting policies in Brazilian Real State Market, as an information predictor for the evaluation process and decision making for long-term projects, comparing the accounting method with the economic one. The essay brings up characteristics of economic environment in this specific segment, presenting the usual business guidelines. In a second moment, a brief study from historical facts of national and international accounting methodologies and their consequences are placed in Brazilian Real State, according to international standards. The financial statement analysis are presented by the optic of financial performance ratios and, in a subsequent section, the methodology for performance analysis is brought out, focused on two different methods for measure the opportunity cost . The differences between accounting and economic profits are pointed. In the end, a case study compares the results.