My thesis is entitled "Contributions to Simulation-based High-dimensional Sequential Decision Making". The context of the thesis is about games, planning and Markov Decision Processes. An agent interacts with its environment by successively making decisions. The agent starts from an initial state until a final state in which the agent can not make decision anymore. At each timestep, the agent receives an observation of the state of the environment. From this observation and its knowledge, the agent makes a decision which modifies the state of the environment. Then, the agent receives a reward and a new observation. The goal is to maximize the sum of rewards obtained during a simulation from an initial state to a final state. The policy of the agent is the function which, from the history of observations, returns a decision. We work in a context where (i) the number of states is huge, (ii) reward carries little information, (iii) the probability to reach quickly a good final state is weak and (iv) prior knowledge is either nonexistent or hardly exploitable. Both applications described in this thesis present these constraints : the game of Go and a 3D simulator of the european project MASH (Massive Sets of Heuristics). In order to take a satisfying decision in this context, several solutions are brought : 1. Simulating with the compromise exploration/exploitation (MCTS) 2. Reducing the complexity by local solving (GoldenEye) 3. Building a policy which improves itself (RBGP) 4. Learning prior knowledge (CluVo+GMCTS) Monte-Carlo Tree Search (MCTS) is the state of the art for the game of Go. From a model of the environment, MCTS builds incrementally and asymetrically a tree of possible futures by performing Monte-Carlo simulations. The tree starts from the current observation of the agent. The agent switches between the exploration of the model and the exploitation of decisions which statistically give a good cumulative reward. We discuss 2 ways for improving MCTS : the parallelization and the addition of prior knowledge. The parallelization does not solve some weaknesses of MCTS; in particular some local problems remain challenges. We propose an algorithm (GoldenEye) which is composed of 2 parts : detection of a local problem and then its resolution. The algorithm of resolution reuses some concepts of MCTS and it solves difficult problems of a classical database. The addition of prior knowledge by hand is laborious and boring. We propose a method called Racing-based Genetic Programming (RBGP) in order to add automatically prior knowledge. The strong point is that RBGP rigorously validates the addition of a prior knowledge and RBGP can be used for building a policy (instead of only optimizing an algorithm). In some applications such as MASH, simulations are too expensive in time and there is no prior knowledge and no model of the environment; therefore Monte-Carlo Tree Search can not be used. So that MCTS becomes usable in this context, we propose a method for learning prior knowledge (CluVo). Then we use pieces of prior knowledge for improving the rapidity of learning of the agent and for building a model, too. We use from this model an adapted version of Monte-Carlo Tree Search (GMCTS). This method solves difficult problems of MASH and gives good results in an application to a word game.

The development and implementation challenges associated with a proposed load allocation paradigm for fault risk assessment and system health management based on uncertain fault diagnostic and failure prognostic information are investigated. Health management actions are formulated in terms of a value associated with improving system reliability, and a cost associated with inducing deviations from a system's nominal performance. Three simulated case study systems are considered to highlight some of the fundamental challenges of formulating and solving an optimization on the space of available supervisory control actions in the described health management architecture. Repeated simulation studies on the three case-study systems are used to illustrate an empirical approach for tuning the conservatism of health management policies by way of adjusting risk assessment metrics in the proposed health management paradigm. The implementation and testing of a real-world prognostic system is presented to illustrate model development challenges not directly addressed in the analysis of the simulated case study systems. Real-time battery charge depletion prediction for a small unmanned aerial vehicle is considered in the real-world case study. An architecture for offline testing of prognostics and decision making algorithms is explained to facilitate empirical tuning of risk assessment metrics and health management policies, as was demonstrated for the three simulated case study systems.

Stochastic control problems that arise in sequential decision making applications typically assume that information used for decision-making is obtained according to a predetermined sampling schedule. In many real applications however, there is a high sampling cost associated with collecting such data. It is therefore of equal importance to determine when information should be collected as it is to decide how this information should be utilized for optimal decision-making. This type of joint optimization has been a long-standing problem in the operations research literature, and very few results regarding the structure of the optimal sampling and control policy have been published. In this thesis, the joint optimization of sampling and control is studied in the context of maintenance optimization. New theoretical results characterizing the structure of the optimal policy are established, which have practical interpretation and give new insight into the value of condition-based maintenance programs in life-cycle asset management. Applications in other areas such as healthcare decision-making and statistical process control are discussed. Statistical parameter estimation results are also developed with illustrative real-world numerical examples.

Stochastic control problems that arise in sequential decision making applications typically assume that information used for decision-making is obtained according to a predetermined sampling schedule. In many real applications however, there is a high sampling cost associated with collecting such data. It is therefore of equal importance to determine when information should be collected as it is to decide how this information should be utilized for optimal decision-making. This type of joint optimization has been a long-standing problem in the operations research literature, and very few results regarding the structure of the optimal sampling and control policy have been published. In this thesis, the joint optimization of sampling and control is studied in the context of maintenance optimization. New theoretical results characterizing the structure of the optimal policy are established, which have practical interpretation and give new insight into the value of condition-based maintenance programs in life-cycle asset management. Applications in other areas such as healthcare decision-making and statistical process control are discussed. Statistical parameter estimation results are also developed with illustrative real-world numerical examples.

O presente trabalho teve como objetivo principal descrever o processo decisório de compra de maçã na cidade de Porto Alegre - RS, buscando aprofundar o conhecimento acerca dos fatores que influenciam a decisão de compra da fruta pelo último elo da cadeia produtiva: o consumidor final. Para a efetiva condução deste trabalho, o processo metodológico apresentou uma pesquisa exploratória, realizada em uma fase qualitativa, em que foram apontados os indicadores relacionados à tomada de decisão de compra do produto, baseados em material pré-existente sobre o tema, e entrevistas de caráter exploratório com especialistas e com consumidores de maçã; e outra fase quantitativa, momento no qual foi aplicado o instrumento de coleta ao público-alvo. Após compreender a situação atual da cadeia produtiva da fruta e concluir a aplicação dos questionários, os dados foram analisados e interpretados através do software estatístico SPSS Versão 16. Os resultados observados possibilitam a compreensão acerca do processo decisório de compra de maçã na cidade de Porto Alegre – RS, e demonstram que os aspectos que motivam o consumo da fruta estão relacionados ao apelo à saúde, sendo que o principal fator decisório no momento da compra está vinculado à qualidade da aparência estética do produto, e que o fator preço não interfere no aumento ou redução do consumo. / Submitted by Marcelo Teixeira (mvteixeira@ucs.br) on 2014-05-13T12:37:26Z No. of bitstreams: 1 Dissertacao Bianca Rockenbach.pdf: 1812003 bytes, checksum: a7cb2853e6f86a2b450f463dd64a39a0 (MD5) / Made available in DSpace on 2014-05-13T12:37:27Z (GMT). No. of bitstreams: 1 Dissertacao Bianca Rockenbach.pdf: 1812003 bytes, checksum: a7cb2853e6f86a2b450f463dd64a39a0 (MD5) / This study aimed to describe the decision making process of buying apple in Porto Alegre - RS, seeking to deepen the knowledge about the factors that influence the decision to buy fruit at last link in the supply chain of Apple: the end consumer. For the effective conduct of this study, the methodological process presented an exploratory study, conducted in a qualitative phase, which were expressed in the indicators related to decision-making product purchase based on pre-existing material on the topic, and exploratory interviews with experts and consumers with apple , and other quantitative phase, at which time the instrument was used to collect the target audience. After understanding the current status of the production chain of fruit and completing the questionnaires, the data were analyzed and interpreted using the statistical software SPSS Version 16. Results observed to enable understanding of the decision making process of buying apple in Porto Alegre - RS, and demonstrate that the aspects that motivate consumption of fruit are related to appeal to health, and the main decision factor when buying is linked to the quality of the aesthetic appearance of the product, and the price did not interfere with the increase or decrease in consumption.

Esta dissertação de mestrado apresenta um novo esquema de controle de taxa hierárquico – HRC – para o padrão MVC – extensão para vídeos de múltiplas vistas do padrão H.264 – com objetivo de melhorar o aproveitamento da largura de banda oferecida por um canal entregando o vídeo comprimido com a melhor qualidade possível. Este esquema de controle de taxa hierárquico foi concebido para controlar de forma conjunta os níveis de quadro e de unidades básicas (BU). O esquema proposto explora a correlação existente entre as distribuições das taxas de bits em quadros vizinhos para predizer de forma eficiente o comportamento dos futuras bitrates através da aplicação de um controle preditivo baseado em modelos – MPC – que define uma ação de controle apropriada sobre as ações de adaptação do parâmetro de quantização (QP). Para prover um ajuste em granularidade fina, o QP é adicionalmente adaptado internamente para cada quadro por um processo de decisão de Markov (MDP) implementado em nível de BU capaz de considerar mapas com Regiões de Interesse (RoI). Um retorno acoplado aos dois níveis supracitados é realizado para garantir a consistência do sistema. Aprendizagem por Reforço é utilizada para atualizar os parâmetros do Controle Preditivo baseado em Modelos e do processo de decisão de Markov. Resultados experimentais mostram a superioridade da utilização do esquema de controle proposto, comparado às soluções estado-da-arte, tanto em termos de precisão na alocação de bits quanto na otimização da razão taxa-distorção, entregando um vídeo de maior qualidade visual nos níveis de quadros e de BUs. / This master thesis presents a novel Hierarchical Rate Control – HRC – for the Multiview Video Coding standard targeting an increased bandwidth usage and high video quality. The HRC is designed to jointly address the rate control at both framelevel and Basic Unit (BU)-level. This scheme is able to exploit the bitrate distribution correlation with neighboring frames to efficiently predict the future bitrate behavior by employing a Model Predictive Control that defines a proper control action through QP (Quantization Parameter) adaptation. To provide a fine-grained tuning, the QP is further adapted within each frame by a Markov Decision Process implemented at BU-level able to take into consideration a map of the Regions of Interest. A coupled frame/BU-level feedback is performed in order to guarantee the system consistency. A Reinforcement Learning method is responsible for updating the Model Predictive Control and the Markov Decision Process parameters. Experimental results show the superiority of the Hierarchical Rate Control compared to state-of-the-art solutions, in terms of bitrate allocation accuracy and rate-distortion, while delivering smooth video quality at both frame and Basic Unit levels.

Submitted by Paulo Junior (paulo.jr@fgv.br) on 2010-12-13T21:15:39Z No. of bitstreams: 1 Carla Winter.pdf: 1507257 bytes, checksum: 938194a6dc53572559e3bd377245140d (MD5) / Approved for entry into archive by Paulo Junior(paulo.jr@fgv.br) on 2010-12-13T21:15:59Z (GMT) No. of bitstreams: 1 Carla Winter.pdf: 1507257 bytes, checksum: 938194a6dc53572559e3bd377245140d (MD5) / Made available in DSpace on 2010-12-16T17:40:10Z (GMT). No. of bitstreams: 1 Carla Winter.pdf: 1507257 bytes, checksum: 938194a6dc53572559e3bd377245140d (MD5) Previous issue date: 2010-03-26 / Within the literature of Consumer Behavior and Decision Theory there is a considerable body of theory that analyzes negative feelings and negative reactions in the decision making process, concerning high and low involvement products. Several phenomena are identified as negative in the process, such as Consumer Confusion, which comprises three dimensions: i) too much similar information about products, ii) a lot of information about different products and iii) false and ambiguous information. This phenomenon, however, seems to be moderated by a set of variables, such as Involvement, Experience and Time Constraints (moderating the relationship between Consumer Confusion and Purchase Intention). This was identified through in-depth interviews. The results of the interviews enabled the identification of moderating variables, as well as the existence of the phenomenon and its relation to the Purchase Intention. In the second phase of the research, it is assumed that individuals with low Involvement and Time Constraints have a greater propensity to confusion. In Study 2, Involvement and Time Constraint were used as moderators, both manipulated by instruction, and the dependent variables were Purchase Intention and Consumer Confusion. The results of Study 2 allowed the reseracher to infer that there are, indeed, significant differences between the groups, when analyzing the variable Consumer Confusion, but in most groups the Purchase Intention was not significantly different. In Study 3, Experience (strong and weak) and Consumer Confusion were manipulated and the dependent variable was Purchase Intention. The results of Study 3 also allowed to infer that there are significant differences between groups referring to Purchase Intention, when considered low or high confusion as well as strong or weak Experience. In the last phase of the research, consumer strategies for dealing with the phenomenon were the focus of the study. Such strategies often mediate subsequent behavior, for instance, purchase itself. In Study 4, Consumer Confusion was manipulated in two of its dimensions: overload and similarity. It was possible to identifie the preponderance of two strategies: search for additional information and postponement of the purchase decision. / Dentro da literatura de Comportamento do Consumidor e Teoria da Decisão existe considerável corpo teórico que analisa sentimentos negativos e reações adversas no processo decisório de compras de produtos de alto e baixo envolvimento. Vários fenômenos são identificados como negativos no processo, principalmente a Confusão do Consumidor, que compreende três dimensões: i) muitas informações similares sobre produtos, ii) muitas informações sobre diferentes produtos e iii) informações falsas e ambíguas. Tal fenômeno, no entanto, parece ser moderado por um conjunto de variáveis, como o Envolvimento, a Experiência e a Restrição de Tempo (moderadoras da relação entre Confusão do Consumidor e Intenção de Compra). Este fato foi identificado através de entrevistas em profundidade. Os resultados das entrevistas permitiram identificar as variáveis moderadoras, assim como a existência do fenômeno e sua relação com a decisão final de compra. Na segunda fase da pesquisa, supõe-se que indivíduos com baixo Envolvimento e Restrição de Tempo possuam uma propensão maior à confusão. No Estudo 2 foram utilizados como moderadores o Envolvimento e a Restrição de Tempo, ambos manipulados por instrução, sendo as variáveis dependentes a Intenção de Compra e a Confusão do Consumidor. Os resultados do Estudo 2 permitiram inferir que existem diferenças significativas entre os grupos, quando analisada a variável Confusão do Consumidor, mas, em alguns grupos, a Intenção de Compra não era significativamente diferente. No Estudo 3 foram manipuladas a Experiência (forte e fraca) e a Confusão do Consumidor, sendo a variável dependente a Intenção de Compra. Os resultados do Estudo 3 também permitiram inferir que existem diferenças significativas entre os grupos na Intenção de compra, quando consideradas baixa ou alta confusão, assim como Experiência forte ou fraca. Na última fase da pesquisa foram destacadas as estratégias dos consumidores para lidar com o fenômeno Confusão do Consumidor. Tais estratégias, muitas vezes, são mediadoras de comportamentos posteriores, como a compra do produto. No Estudo 4 manipulou-se a Confusão do Consumidor em duas de suas dimensões. Foi possível destacar a preponderância da estratégia por busca de informações e postergação da decisão, quando o consumidor se depara com situações confusas.

Nous étudions les systèmes de véhicules en libre service en aller-simple : avec emprunt et restitution dans des lieux éventuellement différents. La publicité promeut l'image de flexibilité et d'accessibilité (tarifaire) de tels systèmes, mais en réalité il arrive qu'il n'y ait pas de véhicule disponible au départ, voire pire, pas de place à l'arrivée. Il est envisageable (et pratiqué pour Vélib' à Paris) de relocaliser les véhicules pour éviter que certaines stations soient vides ou pleines à cause des marées ou de la gravitation. Notre parti-pris est cependant de ne pas considérer de ``relocalisation physique'' (à base de tournées de camions) en raison du coût, du trafic et de la pollution occasionnées (surtout pour des systèmes de voitures, comme Autolib' à Paris). La question à laquelle nous désirons répondre dans cette thèse est la suivante : Une gestion via des tarifs incitatifs permet-elle d'améliorer significativement les performances des systèmes de véhicules en libre service ? / One way Vehicle Sharing Systems (VSS), in which users pick-up and return a vehicle in different places is a new type of transportation system that presents many advantages. However, even if advertising promotes an image of flexibility and price accessibility, in reality customers might not find a vehicle at the original station (which may be considered as an infinite price), or worse, a parking spot at destination. Since the first Bike Sharing Systems (BSS), problems of vehicles and parking spots availability have appeared crucial. We define the system performance as the number of trips sold (to be maximized). BSS performance is currently improved by vehicle relocation with trucks. Our scope is to focus on self regulating systems through pricing incentives, avoiding physical station balancing. The question we are investigating in this thesis is the following: Can a management of the incentives increases significantly the performance of the vehicle sharing systems?

Made available in DSpace on 2016-12-01T19:11:33Z (GMT). No. of bitstreams: 1 120337.pdf: 2530428 bytes, checksum: 1257bde6d7e3a80b40b7a94f981eaa7a (MD5) Previous issue date: 2014-09-24 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The scant bibliography of sports marketing; the practical obstacles encountered in the field of sports management in Brazil; the sport s role in the development of society and its special moment in the country; and the specific case of running races as social&cultural phenomenon they all motivate the conducting of a study that can contribute with more professional sports management actions, oriented to the consumers. The number of running events has grown sharply in recent years, and the trend is to continue growing and developing. The increased competitiveness suggests that understading the consumer s decision process is important to the organizers, so they can promote events more synchronized with consumer s desires. This study aims to identify and assess the attributes of running events that influence the participants decision process. The research is fundamentally descriptive, with exploratory characteristics. The questionary is developed in the following steps: participant-observation of running events; discussion with focus group; and literature related to sports marketing, motivation for participating in sports events and event management. The instrument for data collection characterize the runner considering social and demographical characteristics, connection status with the sport, level of interest and participation in running events, and motivation to participate in these events. The sample is held for convenience and formed by 496 runners. The study identifies that the majority of Brazilian runners are classified in the attachement stage, characterized by a strong emotional connection and the self-identification as a runner; that these runners are motivated to participate in running events specially by two dimensions fitness and skill mastery; and that 58 attributes influence the decision process to participate in running events, being distance, date and price the three most influent. The 58 attributes are submitted to factor analysis, resulting in nine factors: safety, entertainment, physical characteristics, registration process, accessibility, information, reward, differentials, and basics. Significant differences of the influence generated by these attributes in the decision process are found when ANOVA is performed considering the division of runners in groups such as age, number of events concluded by year, connection stage, and degree of interest for various running events. Knowing the specifics of the behavior of diverse running groups is critical to ensure the organization of events that meet the consumers desires. / A escassa bibliografia da área de marketing esportivo no Brasil, as dificuldades práticas observadas no campo da gestão esportiva, o papel do esporte no desenvolvimento da sociedade e seu momento no país, e o caso específico das corridas pedestres como fenômeno sociocultural contemporâneo, motivam a realização de estudo que contribua com uma gestão esportiva mais profissional e focada no consumidor. Os eventos de corrida pedestre cresceram de maneira acentuada nos últimos anos, e a tendência é que continuem se desenvolvendo. A maior competitividade sugere que conhecer o processo de decisão de participação é importante para que organizadores promovam eventos mais sintonizados com os desejos dos consumidores. O objetivo desta pesquisa consiste em identificar e avaliar os atributos de eventos de corrida pedestre que influenciam o processo de decisão de participação nesses eventos. O estudo é fundamentalmente descritivo, com características exploratórias. As perguntas para a elaboração do questionário são desenvolvidas em etapas: observação participante de eventos de corrida pedestre; discussão com grupo focal; e revisão da literatura de marketing esportivo, motivação para participação em eventos esportivos e gerenciamento de eventos. O instrumento de coleta caracteriza o corredor pedestre quanto a aspectos sociodemográficos, nível de participação e interesse nos eventos de corrida pedestre, estado de ligação com a corrida pedestre, e motivação para participação nos eventos. A amostra é realizada por conveniência e formada por 496 indivíduos. Identifica-se que a maior parte dos corredores se encontra no estado de ligação com a corrida, caracterizado por forte componente emocional e auto-identificação do indivíduo como corredor; que esses corredores são motivados especialmente por duas dimensões, a melhora da condição física e o domínio das habilidades de corrida; e que 58 atributos causam influência no processo de decisão de participação, sendo a distância do evento, o preço e a data de realização do mesmo os mais influentes. Os 58 atributos são submetidos à análise fatorial, que resulta em nove fatores: segurança, entretenimento, características físicas, processo de inscrição, acessibilidade, informações, recompensa, diferenciais e básicos. Diferenças significativas da influência desses atributos no processo de decisão são encontradas quando se realiza análise de variância, considerando divisão dos corredores em grupos de idade, número de eventos concluídos por ano, estágio de ligação, e interesse por diferentes eventos. Conhecer as particularidades de comportamento dos diversos grupos de corredores é fundamental para proporcionar eventos orientados a garantir a satisfação dos desejos dos consumidores.

Made available in DSpace on 2015-05-14T12:20:13Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 927858 bytes, checksum: cc8a86dc4e128757eb5249df0f117366 (MD5) Previous issue date: 2014-09-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Analyzes the process of search and use of information by managers of top management subsidiary bodies of the UFPB, and how they use the information to create meanings in decision making. The work is grounded in theory, the organizational model of knowledge developed by Chun Wei Choo, above all, in their quest for perspective and use of information for the creation of meaning. It features forming agents of the organizational environment and ways to access, store, retrieve, process and disseminate information, identifying the behavioral aspects and examining the conditions of the situational environment in the construction of meaning and decision making. The empirical field is formed by academic and administrative pro-rectories UFPB. The subjects are the seven pro-rectors and twenty-two coordinators of the surveyed units, actors interacting in the institutional environment in search and use information. As a research strategy, we adopted a quantitative-qualitative approach, defined by a single case study, using questionnaire to collect data and information. At the end of the study revealed that the process of search and use of information by managers and how they use the information to create meanings are gaps and informational limitations. Managers do not dare to take risks to act in a context of complexity, ambiguity and uncertainty, such as the university. The default behavior, methodical and systematic, minimizes the importance of tacit knowledge, is oblivious to the cognitive, emotional and situational skills and is characterized by conservatism, limited the choices of information sources, as well as the difficulty to perceive and define their own needs information, and a practice-based activities, already institutionalized, search and use of information. The managers of the central government UFPB are conditioned to emergency contingencies of their work unit, because the environmental context recognition processes are not well defined, and the institution does not have them, too, clearly defined and structured. / Analisa-se o processo de busca e uso da informação, por parte dos gestores dos órgãos auxiliares de direção superior da UFPB, e a forma como eles utilizam a informação para criar significados no processo decisório. O trabalho é fundamentado, teoricamente, no modelo de organização do conhecimento desenvolvido por Chun Wei Choo, sobretudo, em sua perspectiva de busca e uso da informação para a criação de significados. Caracteriza os agentes formadores do ambiente organizacional e suas formas de acessar, armazenar, recuperar, tratar e disseminar a informação, identificando os aspectos comportamentais e examinando as condições do ambiente situacional no processo de construção de significados e na tomada de decisão. O campo empírico é formado pelas pró-reitorias acadêmicoadministrativas da UFPB. Os sujeitos da pesquisa são os sete pró-reitores e vinte e dois coordenadores das unidades pesquisadas, atores que interagem no ambiente institucional na busca e no uso de informações. Como estratégia de pesquisa, adotou-se uma abordagem quanti-qualitativa, definida por estudo de caso único, com utilização de questionário na coleta de dados e de informações. Ao final do estudo, ficou evidente que o processo de busca e uso de informação pelos gestores e a forma como utilizam a informação para criar significados apresentam lacunas e limitações informacionais. Os gestores não ousam assumir riscos ao atuar num contexto de complexidade, ambiguidade e incerteza, como o da universidade. O padrão de comportamento, metódico e sistemático, minimiza a importância do conhecimento tácito, é alheio às capacidades cognitivas, emocionais e situacionais e se caracteriza pelo conservadorismo, limitado nas escolhas das fontes de informação, assim como pela dificuldade de perceber e definir as próprias necessidades de informação, e por uma atuação baseada em práticas, já institucionalizadas, de busca e uso da informação. Os gestores da administração central da UFPB são condicionados às contingências emergenciais da sua unidade de trabalho, porquanto os processos de reconhecimento do contexto ambiental não são bem definidos, e a instituição não os tem, também, claramente definidos e estruturados.