Dissertationen zum Thema „Logit“

This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Thesis: S.M. in Transportation, Massachusetts Institute of Technology, Department of Civil and Environmental Engineering, 2019
Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 67-68).
This work is about developing an estimation procedure for nested logit models that optimizes over the nesting structure in addition to the model parameters. Current estimation practices require an a priori specification of a nesting structure. We formulate the problem of learning an optimal nesting structure as a mixed integer nonlinear programming (MINLP) optimization problem and solve it using a variant of the linear outer approximation algorithm. We demonstrate that it is indeed possible to recover the nesting structure directly from the data by applying our method to synthetic and real datasets.
by Youssef Medhat Aboutaleb.
S.M. in Transportation
S.M.
S.M.inTransportation Massachusetts Institute of Technology, Department of Civil and Environmental Engineering
S.M. Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science

The effects of random guessing as a measurement disturbance on Rasch fit statistics (unweighted total, weighted total, and unweighted ability between) and the Logit Residual Index (LRI) were examined through simulated data sets of varying sample sizes, test lengths, and distribution types. Three test lengths (25, 50, and 100), three sample sizes (25, 50, and 100), two item difficulty distributions (normal and uniform), and three levels of guessing (no guessing (0%), 25%, and 50%) were used in the simulations, resulting in 54 experimental conditions. The mean logit person ability for each experiment was +1. Each experimental condition was simulated once in an effort to approximate what could happen on the single administration of a four option per item multiple choice test to a group of relatively high ability persons. Previous research has shown that varying item and person parameters have no effect on Rasch fit statistics. Consequently, these parameters were used in the present study to establish realistic test conditions, but were not interpreted as effect factors in determining the results of this study.

Eine Klasse von statistischen Tests zur Spezifikation von Logit-Modellen wird vorgestellt. Diese Tests sind Adaptionen von allgemeineren Spezifikationstests nichtlinearer Modelle. Die theoretische Anwendbarkeit der Tests wird dargelegt und die praktische Anwendbarkeit in Simulationsstudien erörtert. Eine konkrete Anwendung auf einen Datensatz und ein multinomiales Logit-Modell zur Produktwahl wird präsentiert.

The multinomial logit model (MNL) possesses a latent variable representation in terms of random variables following a multivariate logistic distribution. Based on multivariate finite mixture approximations of the multivariate logistic distribution, various data-augmented Metropolis-Hastings algorithms are developed for a Bayesian inference of the MNL model.

Gene expression studies are of growing importance in the field of medicine. In fact, subtypes within the same disease have been shown to have differing gene expression profiles (Golub et al., 1999). Often, researchers are interested in differentiating a disease by a categorical classification indicative of disease progression. For example, it may be of interest to identify genes that are associated with progression and to accurately predict the state of progression using gene expression data. One challenge when modeling microarray gene expression data is that there are more genes (variables) than there are observations. In addition, the genes usually demonstrate a complex variance-covariance structure. Therefore, modeling a categorical variable reflecting disease progression using gene expression data presents the need for methods capable of handling an ordinal outcome in the presence of a high dimensional covariate space. In this research we present a method that combines the stereotype regression model (Anderson, 1984) with an elastic net penalty (Friedman et al., 2010) as a method capable of modeling an ordinal outcome for high-throughput genomic datasets. Results from applying the proposed method to both simulated and gene expression data will be reported and the effectiveness of the proposed method compared to a univariable and heuristic approach will be discussed.

Nessa dissertação, procura-se avaliar se o retorno esperado do ensino superior é determinante para a escolha profissional e explicar como os salários, as habilidades e as características sócio-econômicas dos alunos da Universidade de São Paulo (USP), em conjunto, se associam à escolha da carreira. Para atingir esses objetivos, são utilizados dados da Fundação Universitária para o Vestibular (Fuvest) de 1995 e de 1996 e do Censo 2000. As carreiras são divididas em seis áreas de atuação – educação, ciências humanas, negócios, saúde, engenharia e matemática/ciências – e são estimados dois modelos de escolha discreta: logit multinomial e logit condicional. Na estimação do logit multinomial são utilizadas apenas variáveis relativas às características dos indivíduos e no logit condicional inclui-se o salário médio de cada área de atuação. Os resultados indicam que o retorno esperado não tem efeito sobre a escolha profissional.
This research evaluate if higher education expected return is determinant for the professional choice and how the interaction between wage, abilities and socio-economic characteristics of Universidade de São Paulo’s students, all together, are associated to the career choice. Data used is from Fundação Universitária para o Vestibular (Fuvest) 1995 and 1996’s questionnaires and from Censo 2000. The careers are grouped in six concentration areas – education, human sciences, management, health, engineer, math/sciences – and two discrete choice models are estimated: multinomial logit and conditional logit. Multinomial logit contains only variables specific to individuals and the average wage of each concentration area is included in conditional logit estimation. The results show that expected return doesn’t have effect over the professional choice.

Thesis (M. S.)--Industrial and Systems Engineering, Georgia Institute of Technology, 2008.
Committee Chair: Anton Kleywegt; Committee Co-Chair: Alexander Shapiro; Committee Member: Charles Rosa; Committee Member: Shabbir Ahmed; Committee Member: Sigrun Andradottir.

Divers types de modeles logit et probit sont etudies en vue de l'explication des modalites prises par une ou plusieurs variables, en fonction de variables explicatives. Dans le cadre d'une seule variable binaire, le modele dichotomique simple logit et probit ainsi que ses erreurs de specification sont traites. Les modeles logit et probit polytome multinomial ou conditionnel sont analyses lorsque la variable a expliquer est a plusieurs modalites. L'etude et la modelisation simultanee de plusieurs variables qualitatives fait appel au modele descriptif et explicatif log-lineaire, au modele a reponses multinomiales, ou aux modeles a equations simultanees logit et probit lorsque parmi les variables explicatives figurent des variables endogenes. Enfin dans le cas de donnees binaires individuelles temporelles et sous differentes hypotheses de vraie dependance entre les etats du processus, de non stationnarite et d'heterogeneite, nous etudions les solutions theoriques et la mise en oeuvre informatique des modeles logit et probit associes
Several types of logit and probit models are studied in ordrer to explain levels related to one or many qualitative variables by explanatory variables. Simple dichotomous logit and probit models and their specification errors are discussed. Are also analyzed multinomial and conditional polychotomous in the case of logit and probit models when the dependent variable possesses several levels. The study and the simultaneous modelisation of two or many qualitative variables deal with descriptive and explanatory log linear model, multinomial responses models, and simultaneous equations logit and probit models when some of the explanatory variables are endogeneous. Finally in the case of individual temporal binary data and with some different true dependency hypothesis between process states, non stationarity and heterogeneity, we study not only the theorical solutions but the computational work of logit and probit models associated as well

Trip generation is the first stage of the conventional 'four-stage' transport model. The aim of this stage is to predict total number of trips generated to and from each zone. The two most common techniques for trip generation are linear regression (the dependent vaziable is alinear-in-parameter function of a number of explanatory variables) and category analysis including multiple classification analysis (based on estimating number of trip generations as a function of household attributes). Both techniques of trip generation rely on the availability of a large socio-economic, mainly revealed preference data set. They also have technical limitations such as the assumption of linearity which might result in unreasonable predictions of trip generation. Any deficiency or inaccuracy in the estimation at this stage will be carried over and will have implications on subsequent stages. The other stages of the 'four-stage' model employ other techniques including logistic analysis which broadens the scope of the analysis. Logistic regression analysis has been used to model travel choices such as mode, route and departure time but not trip generation. There has not been much research to investigate the appropriateness of using this technique to model generation. The main reason for this is that logistic regression predicts probabilities rather than the total number of trips. In order to be able to model trip generation using logistic regression, the number of trips frequency) can be treated as a set of mutually exclusive categorical variables; therefore the built-in upper and lower limits are incorporated. Therefore, it is not possible to predict a negative number of trips and the estimates of the model will show the underlying probabilities for the actual number of This will also provide a behavioural framework that directly links the number of trips to utility-based consumer and decision-making theory. Logistic regression can be used to model trip generation as binary, multinomial or nested logit frameworks. An added advantage of using this approach is the ability to predict the frequency and number of trips made by each individual. The aim of this research therefore, is to investigate possible methodologies to improve performance of trip generation modelling. In order to achieve this aim firstly, this research investigates the appropriateness of logistic regression to model trip generation and device a methodology for it. The analysis and comparisons of the results with results from conventional models are examined. Exploring the use of stated preference data to calibrate trip generation models is also studied here. Finally, transport policy measures and enhanced transport accessibility functions have been investigated in generation models.

Made available in DSpace on 2014-06-12T17:20:04Z (GMT). No. of bitstreams: 2 arquivo573_1.pdf: 3504462 bytes, checksum: 028fb584698b08755667d8902b92ee48 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2010
Conselho Nacional de Desenvolvimento Científico e Tecnológico
Este trabalho teve como objetivo principal analisar como os estudantes candidatos às vagas na Universidade Federal de Pernambuco tem escolhido qual carreira seguir. Desta forma foi estudado como as características sócio-econômicas dos candidatos influenciam na escolha e, em seguida, foi analisado como os retornos salariais correspondentes ao grupos de cursos disponíveis se associavam às características individuais para influenciar na decisão de qual profissão seguir. Para isto, foram utilizados os dados da Covest, que é a comissão organizadora do vestibular da Universidade Federal de Pernambuco, para o vestibular 2009. Além disso também foram utilizados dados da PNAD 2008 para a estimação dos salário médio de cada carreira. Os cursos oferecidos pela UFPE foram divididos em 9 grupos, obedecendo a razões de proximidade profissional das carreiras, e para realizar as estimações foram utilizados os modelos econométricos de logit multinomial e logit condicional. O primeiro modelo utiliza apenas variáveis explicativas relacionadas aos indivíduos e o segundo é quando se utiliza alguma variável explicativa característica das alternativas. Os resultados apontam para a influência que as variáveis pessoais, background familiar e variáveis educacionais exercem sobre a probabilidade de escolha dos grupos de cursos disponíveis. Também foi constatado, que o retorno salarial não tem significância sobre a probabilidade de escolha quando se leva em conta as variáveis que representam as características individuais

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-09-05T15:05:52Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Kelly_Samá_Lopes_de_Vasconcelos.pdf: 899498 bytes, checksum: 7b70712927d11b795cc4e48b29a641d6 (MD5)
Made available in DSpace on 2016-09-05T15:05:52Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Kelly_Samá_Lopes_de_Vasconcelos.pdf: 899498 bytes, checksum: 7b70712927d11b795cc4e48b29a641d6 (MD5) Previous issue date: 2016-04-29
CAPEs
PROPESQ
A informalidade tem sido constantemente discutida na literatura econômica, e os resultados sugerem que uma vez que não existe recolhimento de tributos, são gerados impactos na arrecadação, sendo a Previdência a maior prejudicada. O objetivo desta pesquisa é analisar, a partir de características do empresário individual, as chances desses empresários contribuírem com a Previdência ou passarem a ter inscrição no Cadastro Nacional de Pessoa Jurídica, a partir da estimação de um modelo de dados em painel logit. As estimativas obtidas mostraram que as chances de formalização dos empreendedores individuais parece ter sido afetada positivamente de 2009 a 2014 ao considerar a contribuição previdenciária; já para o critério CNPJ, os resultados foram positivos, mas de pequena magnitude. Com relação às mudanças nas chances de contribuição com a Previdência, de 2004 a 2008 para 2009 a 2014, pode-se dizer que houve um aumento de 2009 a 2014 das chances de o empresário individual contribuir com a Previdência. Uma possível explicação para a ocorrência de resultados com baixa magnitude na formalização via inscrição no CNPJ estaria no curto período de análise, ou seja, são efeitos de curto prazo.
Informality has been constantly discussed in the economic literature, and the results suggest that since there is nonpayment of taxes are generated impacts on revenues, and the social security most impaired. In this paper the objective is analyze from individual entrepreneur characteristics, the chances of these entrepreneurs contribution to the security or official tax payer administrative file (CNPJ) by the informal individual entrepreneurs, from the estimation of a logit panel model. The estimates obtained demonstrated the odds of formalization of individual entrepreneurs appears to have been positively affected (2009-2014) to consider the social security contributions, as the criterion for CNPJ, the results are positive, but small in magnitude. One possible explanation for the occurrence of results with low magnitude in the formalization of CNPJ would be the short period of analysis.

Nos ensaios de dose e resposta, geralmente são aplicadas diferentes intensidades (doses) de um estímulo (produto) a diferentes grupos de uma população de indivíduos, que poderão responder ou não. A dose máxima que cada indivíduo tolera é chamada tolerância e no caso de se considerar toda a população, ter-se-á uma distribuição de tolerâncias. Muitas vezes, porém, dentro de uma mesma população, existe uma mistura de indivíduos de diferentes sexos, raças, estágios de desenvolvimento, etc., caracterizando subpopulações com diferentes distribuições de tolerâncias. LWIN & MARTIN (1989) apresentam uma mistura de modelos probit". Neste trabalho, faz-se um estudo geral para misturas de modelos"probit","logit"e"complemento-log-log". Para a estimação dos parâmetros desses modelos, utilizou-se o método de escores de Fisher e uma modificação do algoritmo EM, implementados através dos pacotes computacionais Excel 4.0 for Windows e GLIM. Para exemplificar essa metodologia, foi conduzido e analisado um ensaio de resistência da mosca (Musca. Domestica.) ao inseticida deltametrina. Analisou-se, ainda, um ensaio de sensibilidade de um isolado do fungo Metarhizium anisopliae a diferentes doses de radiação e reanalisou-se o ensaio de resistência dos ovos do verme Ostergagia. Spp. a diferentes doses de tiabendazol (TEZ), apresentado por LWIN & MARTIN (1989). Observou-se a adequação do método aos casos em que havia razões biológicas para a existência de uma mistura de distribuições, confirmada pelo mau ajuste de modelos para uma única distribuição. Confirmando o que afirmaram TITTERINGTON & MORGAN (1977), observou-se, ainda, na estimação dos parâmetros do modelo de misturas, a menor velocidade de convergência do algoritmo EM, quando comparado com o método de Newton-Raphson modificado (método de escores de Fisher)
not available

O objectivo central deste trabalho e avaliar o efeito de se considerarem medições aproximadas das exposições na incidência de doenças resultantes de impactos ambientais. Dado existirem quase exclusivamente estatísticas de incidência das doenças, de- senvolvemos modelos que, a partir dessas estatísticas, permitam abordar o nosso problema. Em particular, mostrámos que, quando se utilizam modelos logit, o uso de medidas aproximadas dos impactos ambientais leva a uma distorção, por defeito, dos coeficientes angulares das rectas ajustadas. Estudámos ainda limites para essa distorção, utilizando para tal a aproximação de Edgeworth. Os nossos resultados permitiram-nos também esquematizar diversos cenários para delineamento de trabalhos de campo, indispensáveis ao aprofundamento do nosso objectivo.

2010 - 2011
A complex system is generally de_ned as a system emerging from the interaction of several and di_erent components, each one with their properties and their goals, usually subject to external inuences. Nowadays, complex systems are ubiquitous and they are found in many research areas: examples can be found in Economy (e.g., markets), Physics (e.g., ideal gases, spin systems), Biology (e.g., evolution of life) and Computer Science (e.g., Internet and social networks). Modeling complex systems, understanding how they evolve and predicting the future status of a complex system are major research endeavors. Historically, physicists, economists, sociologists and biologists have separately studied complex systems, developing their own tools that, however, often are not suitable for being adopted in di_erent areas. Recently, the close relation between phenomena in di_erent research areas has been highlighted. Hence, the aim is to have a powerful tool that is able to give us insight both about Nature and about Society, an universal language spoken both in natural and in social sciences, a modern code of nature. In a recent book [16], Tom Siegfried pointed out game theory as such a powerful tool, able to embrace complex systems in Economics [3, 4, 5], Biology [13], Physics [8], Computer Science [10, 11], Sociology [12] and many other disciplines. Game theory deals with sel_sh agents or players, each with a set of possible actions or strategies. An agent chooses a strategy evaluating her utility or payo_ that does not depend only on agent's own strategy, but also on the strategies played by the other players. The way players update their strategies in response to changes generated by other players de_nes the dynamics of the game and describes how the game evolves. If the game eventually reaches a _xed point, i.e., a state stable under the dynamics considered, then it is said that the game is in an equilibrium, through which we can make predictions about the future status of a game. The classical game theory approach assumes that players have complete knowledge about the game and they are always able to select the strategy that maximizes their utility: in this rational setting, the evolution of a system is modeled by best response dynamics and predictions can be done by looking at well-known Nash equilibrium. Another approach is followed by learning dynamics: here, players are supposed to \learn" how to play in the next rounds by analyzing the history of previous plays. By examining the features and the drawbacks of these dynamics, we can detect the basic requirements to model the evolution of complex systems and to predict their future status. Usually, in these systems, environmental factors can inuence the way each agent selects her own strategy: for example, the temperature and the pressure play a fundamental role in the dynamics of particle systems, whereas the limited computational power is the main inuence in computer and social settings. Moreover, as already pointed by Harsanyi and Selten [9], the complete knowledge assumption can fail due to limited information about external factors that could inuence the game (e.g., if it will rain tomorrow), or about the attitude of other players (if they are risk taking), or about the amount of knowledge available to other players. Equilibria are usually used to make predictions about the future status of a game: for this reason, we like that an equilibrium always exists and that the game converges to it. Moreover, in case that multiple equilibria exist, we like to know which equilibrium will be selected, otherwise we could make wrong predictions. Finally, if the dynamics takes too long time to reach its _xed point, then this equilibrium cannot be taken to describe the state of the players, unless we are willing to wait super-polynomially long transient time. Thus we would like to have dynamics that models bounded rationality and induces an equilibrium that always exists, it is unique and is quickly reached. Logit dynamics, introduced by Blume [6], models a noisy-rational behavior in a clean and tractable way. In the logit dynamics for a game, at each time step, a player is randomly selected for strategy update and the update is performed with respect to an inverse noise parameter _ (that represents the degree of rationality or knowledge) and of the state of the system, that is the strategies currently played by the players. Intuitively, a low value of _ represents the situation where players choose their strategies \nearly at random" because they are subject to strong noise or they have very limited knowledge of the game; instead, an high value of _ represents the situation where players \almost surely" play the best response, that is, they pick the strategies yielding high payo_ with higher probability. This model is similar to the one used by physicists to describe particle systems, where the behavior of each particle is inuenced by temperature: here, low temperature means high rationality and high temperature means low rationality. It is well known [6] that this dynamics de_nes an ergodic _nite Markov chain over the set of strategy pro_les of the game, and thus it is known that a stationary distribution always exists, it is unique and the chain converges to such distribution, independently of the starting pro_le. Since the logit dynamics models bounded rationality in a clean and tractable way, several works have been devoted to this subject. Early works about this dynamics have focused about long-term behavior of the dynamics: Blume [6] showed that, for 2 _ 2 coordination games and potential games, the long-term behavior of the system is concentrated in a speci_c Nash equilibrium; Al_os-Ferrer and Netzer [1] gave a general characterization of long term behavior of logit dynamics for wider classes of games. A lot of works have been devoted to evaluating the time that the dynamics takes to reach speci_c Nash equilibria of a game, called hitting time: Ellison [7] considered logit dynamics for graphical coordination games on cliques and rings; Peyton Young [15] extended this work for more general families of graphs; Montanari and Saberi [14] gave the exact graph theoretic property of the underlying interaction network that characterizes the hitting time in graphical coordination games; Asadpour and Saberi [2] studied the hitting time for a class of congestion games. Our approach is di_erent: indeed, our _rst contribution is to propose the stationary distribution of the logit dynamics Markov chain as a new equilibrium concept in game theory. Our new solution concept, sometimes called logit equilibrium, always exists, it is unique and the game converges to it from any starting point. Instead, previous works only take in account the classical equilibrium concept of Nash equilibrium, that it is known to not satisfying all the requested properties. Moreover, the approach of previous works forces to consider only speci_c values of the rationality parameter, whereas we are interested to analyze the behavior of the system for each value of _. In order to validate the logit equilibrium concept we follow two di_erent lines of research: from one hand we evaluate the performance of a system when it reaches this equilibrium; on the other hand we look for bounds to the time that the dynamics takes to reach this equi- librium, namely the mixing time. This approach is trained on some simple but interesting games, such as 2_2 coordination games, congestion games and two team games (i.e., games where every player has the same utility). Then, we give bounds to the convergence time of the logit dynamics for very interesting classes of games, such as potential games, games with dominant strategies and graphical coordination games. Speci_cally, we prove a twofold behavior of the mixing time: there are games for which it exponentially depends on _, whereas for other games there exists a function independent of _ such that the mixing time is always bounded by this function. Unfortunately, we show also that there are games where the mixing time can be exponential in the number of players. When the mixing is slow, in order to describe the future status of the system through the logit equilibrium, we need to wait a long transient phase. But in this case, it is natural to ask if we can make predictions about the future status of the game even if the equilibrium has not been reached yet. In order to answer this question we introduce the concept of metastable distribution, a probability distribution such that the dynamics quickly reaches it and spends a lot of time therein: we show that there are graphical coordination games where there are some distributions such that for almost every starting pro_le the logit dynamics rapidly converges to one of these distributions and remains close to it for an huge number of steps. In this way, even if the logit equilibrium is no longer a meaningful description of the future status of a game, the metastable distributions resort the predictive power of the logit dynamics. References [1] Carlos Al_os-Ferrer and Nick Netzer. The logit-response dynamics. Games and Economic Behavior, 68(2):413 { 427, 2010. [2] Arash Asadpour and Amin Saberi. On the ine_ciency ratio of stable equilibria in congestion games. In Proc. of the 5th International Workshop on Internet and Network Economics (WINE'09), volume 5929 of Lecture Notes in Computer Science, pages 545{ 552. Springer, 2009. [3] Robert J. Aumann and S. Hart, editors. Handbook of Game Theory with Economic Applications, volume 1. Elsevier, 1992. [4] Robert J. Aumann and S. Hart, editors. Handbook of Game Theory with Economic Applications, volume 2. Elsevier, 1994. [5] Robert J. Aumann and S. Hart, editors. Handbook of Game Theory with Economic Applications, volume 3. Elsevier, 2002. [6] Lawrence E. Blume. The statistical mechanics of strategic interaction. Games and Economic Behavior, 5:387{424, 1993. [7] Glenn Ellison. Learning, local interaction, and coordination. Econometrica, 61(5):1047{ 1071, 1993. [8] Serge Galam and Bernard Walliser. Ising model versus normal form game. Physica A: Statistical Mechanics and its Applications, 389(3):481 { 489, 2010. [9] John C. Harsanyi and Reinhard Selten. A General Theory of Equilibrium Selection in Games. MIT Press, 1988. [10] Elias Koutsoupias and Christos H. Papadimitriou. Worst-case equilibria. Computer Science Review, 3(2):65{69, 2009. Preliminary version in STACS 1999. [11] Hagay Levin, Michael Schapira, and Aviv Zohar. Interdomain routing and games. In STOC, pages 57{66, 2008. [12] Jan Lorenz, Heiko Rauhut, Frank Schweitzer, and Dirk Helbing. How social inuence can undermine the wisdom of crowd e_ect. Proceedings of the National Academy of Sciences, 108(22):9020{9025, 2011. [13] John Maynard Smith. Evolution and the theory of games. Cambridge University Press, 1982. [14] Andrea Montanari and Amin Saberi. Convergence to equilibrium in local interaction games. In Proc. of the 50th Annual Symposium on Foundations of Computer Science (FOCS'09). IEEE, 2009. [15] Hobart Peyton Young. The di_usion of innovations in social networks, chapter in \The Economy as a Complex Evolving System", vol. III, Lawrence E. Blume and Steven N. Durlauf, eds. Oxford University Press, 2003. [16] Tom Siegfried. A Beautiful Math: John Nash, Game Theory, and the Modern Quest for a Code of Nature. Joseph Henry Press, 1st ed edition, 2006. [edited by author]
X n.s.

Understanding of the shift behaviours are critical to formulate effective policies to encourage mode shift and anticipate the increasing demands of public transport services impacted by the shifting behaviours. This study aims to investigate the mode choice behaviour of commuters by utilising the nation-wide survey of commuters in Australia and employing the state-of-the-art Multinomial Logit Models. This study has investigated three critical choice behaviours issues: commuters' satisfaction towards train fare, consistency between commuters' current behaviours against their future preferences, and policy interventions to influence mode shift. Its findings have demonstrated significant theoretical and practical contributions to the commuters' choice behaviours.

The purpose of this thesis was to evaluate the impact of dynamic message signs (DMS) on the Orlando-Orange County Expressway Authority (OOCEA) toll road network using the Post-Deployment DMS Survey analysis. DMS are electronic traffic signs used on roadways to give travelers information about travel times, traffic congestion, accidents, disabled vehicles, AMBER alerts, and special events. The particular DMS referred to in this study are large rectangular signs installed over the travel lanes and these are not the portable trailer mount signs. The OOCEA has added twenty-nine fixed DMS to their toll road network from 2006-2008. At the time of the post-deployment survey, a total of twenty-nine DMS were up and running on the OOCEA toll road network. Since most of the travelers on the OOCEA toll roads were from Orange, Osceola, and Seminole counties, this study was limited to these counties. This thesis documents the results for the post-deployment survey analysis. The instrument used to analyze the travelers' perception of DMS was a survey that utilized computer aided telephone interview. The post-deployment survey was conducted during the month of May, 2008. Questions pertaining to the acknowledgement of DMS on the OOCEA toll roads, satisfaction with travel information provided on the network, formatting of the messages, satisfaction with different types of messages, diversion questions (Revealed and Stated preferences), and classification/socioeconomic questions (such as age, education, most traveled toll road, county of residence, and length of residency) were asked to the respondents. This thesis is using results of the multinomial logit model for diversion of traffic. This model takes into account the different diversion decisions from the post development survey (stay vs. divert all the way vs. divert and come back vs. abandon trip) and explains the differences in the diversion behavior. Drivers that use SunPass or Epass tend to stay on the toll road during unexpected congestion. Frequent SR 408 users are more likely to divert and stay off the toll road and frequent SR 417 users are more likely to divert and get back on the toll road. Drivers whose stated preference was to divert off the toll road were more likely to do the same in the real world. However, not too many of the respondents were likely to abandon their trips in the real world even if they said they would in a hypothetical congestion scenario. Users of 511 were more likely to divert and get back on the toll road or abandon their trips due to unexpected congestion. OOCEA can use this study to concentrate on keeping their toll roads more attractive during unexpected congestion to keep drivers from diverting all the way or abandoning their trips. For example, better incident management in clearing accidents more efficiently (thereby decreasing delay) and encouraging the use of SunPass or EPass could help drivers stay than divert or abandon their trip. This thesis also used ordered logit model for satisfaction. This model explains the levels of magnitude of satisfaction with traveler information on OOCEA toll roads. Drivers who acquired traveler information from DMS were less likely to be dissatisfied with traveler information provided on toll roads than other respondents. Drivers who were satisfied with accuracy and information on hazard warnings on DMS were more likely to be satisfied with information provided on toll roads than other respondents. This thesis provides a microscopic insight on the driver behavior on toll roads. This thesis expands the diversion and satisfaction models from previous studies in a way that OOCEA can identify specific groups of drivers related to a given response behavior (i.e., diverts off toll roads or dissatisfied with traveler information). Such analysis can be conducted in the future in the same study area or replicated in other areas to quantify the effects of individual and choice related attributes on choice behavior.
M.S.
Department of Civil and Environmental Engineering
Engineering and Computer Science
Civil Engineering MS

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior
Objetiva-se analisar a pobreza nas Ãreas rurais no Estado do Cearà no que se refere Ãs suas causas socioeconÃmicas como educaÃÃo, idade, cor, gÃnero, atividade principal de ocupaÃÃo e transferÃncias governamentais. AlÃm disso, objetiva-se analisar as caracterÃsticas dos domicÃlios rurais relacionadas ao acesso a bens pÃblicos. Estudam-se as causas da pobreza rural cearense com o objetivo de verificar possÃveis alteraÃÃes no nÃvel da pobreza no perÃodo anterior e posterior do Plano Real. A partir dos microdados das PNAD, obtidos junto ao IBGE, determinam-se as variÃveis que apresentam associaÃÃo com a pobreza rural, por meio do modelo de regressÃo Logit e para analisar os determinantes dos indivÃduos, calculam-se as contribuiÃÃes marginais. Ademais, sÃo calculadas as estatÃsticas descritivas das variÃveis relacionadas Ãs causas da pobreza e ao acesso a bens pÃblicos. Conclui-se que as variÃveis sexo, nÃvel de escolaridade, recebimento de aposentadoria, pensÃo e atividade principal de trabalho possuem relaÃÃo significativa com a pobreza da regiÃo, sendo as principais a educaÃÃo e as transferÃncias governamentais. AlÃm disso, aumentou o acesso aos bens pÃblicos, fato este que proporcionou melhor qualidade de vida para a populaÃÃo rural cearense.
The poverty aims to analyse in the rural areas in the State of the Cearà in which it refers to his causes socioeconomics, like education, age, color, type, principal activity of occupation and government transfers. Besides, there aim to analyse the characteristics of the rural residences made a list to the access the public goods. There are studied the causes of the rural poverty of Cearà with the objective to check possible alterations in the level of the poverty in the previous and subsequent period of the Real Plan. From the microdata of the PNAD, obtained near the IBGE, there are determined the variables that present association with the rural poverty, through the model of regression logit and to analyse the determinants of the individuals, the marginal contributions are calculated. Besides, the public goods are calculated the descriptive statistics of the variables made a list to the causes of the poverty and to the access. It is ended that the variables sex, level of schooling, receiving of retirement, boarding house and principal activity of work have significant relation with the poverty of the region, being the principals to education and the government transfers. Besides, it increased the access to the public goods, this fact that provided better quality of life for the rural population of CearÃ.

Made available in DSpace on 2014-06-12T17:21:58Z (GMT). No. of bitstreams: 2 arquivo6593_1.pdf: 1372457 bytes, checksum: 7d328b5860db2137060c9694c0610e5c (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2011
Conselho Nacional de Desenvolvimento Científico e Tecnológico
Essa dissertação é dividida em dois ensaios. O primeiro objetiva caracterizar a demanda por medicamentos no Brasil. Para isso é feita uma revisão da literatura sobre o tema e expostas suas particularidades para o caso brasileiro por meio dos microdados da POF 2002-2003. Com isso são estimadas as probabilidades de gastos por categoria de medicamento. Conforme os resultados, em geral, essas probabilidades são crescentes com a renda e o grau de instrução do chefe da família. Contudo, outras características como raça, região e idade possuem relevantes efeitos sobre essas probabilidades, onde isso dependerá também do medicamento adquirido. O segundo ensaio visa apresentar como se dá o acesso aos medicamentos pelos brasileiros. Para tanto são aplicados modelos de Escolha discreta, para em seguida, com os dados da PNAD-Saúde 2009, a partir de estimações por Multinomial Logit e Nested Logit, realizar inferências sobre a tomada de decisão de consumo dos indivíduos. Os resultados sugerem que a renda, o grau de instrução e o tipo de doenças infringidas são fatores relevantes. Onde a presença deste último fator afeta diretamente a tomada de decisão de consumo dos brasileiros, principalmente, no caso das doenças mais danosas à saúde

Los efectos o externalidades de red son factores que pueden condicionar las decisiones de contratación de los consumidores en favor de empresas ya establecidas y en contra de los nuevos competidores, pudiendo limitar la competencia efectiva y potencial de los mercados, en especial, en aquellas industrias donde el número de empresas es bajo y la entrada de nuevos competidores es poco frecuente. Por ello, es importante verificar su existencia y la magnitud de sus efectos sobre las decisiones de compra de los consumidores con el objetivo de justificar o establecer medidas que impulsen una competencia más equilibrada entre las empresas. Además, teniendo en consideración que los consumidores pueden tener cierto grado de heterogeneidad en sus comportamientos de adquisición, también resulta relevante estudiar el grado de diferenciación de los efectos de red entre los consumidores a fin de mejorar las políticas que fomenten la competencia. Este trabajo tiene por objetivo estimar un modelo logit mixto bajo el enfoque de la inferencia bayesiana, para estudiar empíricamente la existencia y heterogeneidad de los efectos de red sobre las decisiones de contratación de los consumidores en la industria de telefonía móvil peruana. El análisis se hace con base a una muestra que combina información de la Encuestas Residencial de Servicios de Telecomunicaciones (ERESTEL) del a˜no 2015 e información de las empresas operadoras del servicio de telefonía móvil. Los resultados de las estimaciones realizadas sugieren que los efectos de red tendrían un condicionamiento importante sobre las decisiones de contración del servicio de telefonía móvil, además de presentar un grado de heterogeneidad estadísticamente significativo en la magnitud de sus efectos.
Tesis

Den här uppsatsen inleds med att studera de moment som används för multinomial logistisk regression och hur resultaten mäts. Teorin tar sin avsats i den binomiala logistiska regression, för att stegvis ta sig vidare till den multinomiala logistiska regressionen. Begreppen logit, oddskvoten och sannolikheterna förtydligas, effekterna av de oberoende variablerna diskuteras och kopplingen till vanlig linjär regression åskådliggörs. Det blir även en fördjupning av matematiken bakom den logistiska funktionen. Därefter tillämpas den multinomial logistisk regressionsanalysen med ett praktiskt exempel. Analysmodellen är användbar inom flertalet områden och den här uppsatsen ligger inom ramen för sportanalys. Matchstatistik från ishockey och närmare bestämt Örebro Hockeys matcher från säsongerna 2012/13 till 2017/18 nyttjas och den slutgiltiga modellen använder sig av tre förklarande variabler. Resultatet visar att utfallet efter ordinarie tid kan förklaras till 60,9% med hjälp av matchstatistiken, vilket tyder på att den multinomiala regressionsmodellen presterar likvärdigt med andra metoder som tillämpar kategorisk dataanalys inom sportanalys.
This thesis starts by studying the multinomial logistic regression and its moments and how the results are measured. The theory begins with the binomial logistics regression and gradually moves on towards the multinomial logistics regression. Concepts as logit, odds ratio and probabilities are explained, the effects of the independent variables discussed and the link to ordinary linear regression is illustrated. There will also be a deeper, mathematical look at the function of logistic growth. Thereafter the multinomial logistic regression model will be applied. The model is useful within several domains and this thesis lies within sportsanalytics. For this thesis matchstatistics from ice hockey, that is Örebro Hockey’s matches from season 2012/13 to 2017/18, has been used and the final model has three exploratory variables. The outcome of the result performs equivalent to other methods, which applies categorical data analysis within sportsanalytics.

This thesis consists of two parts. The first part analyzesthe accessibility, generation and license holding effects incar ownership models. The second part develops a route choicemodeling framework with an attempt to address the differencesin drivers' route choice behavior. These two parts of work areboth based on the discrete choice theory - the car ownershipmodels are built up on the standard logit model, whereas theroute choice models are formulated in a mixed logit form.

The study result of the first part shows that measuring theaccessibility by the monetary inclusive value reasonably wellcaptures the mechanism of the accessibility impact. Otheraccessibility proxies such as the parking costs, parking typeand house type are correlated with the accessibility but not toa great extent. Both young and old households are less likelyto have a car. The reduction of the propensity to own a car issignificant for households with average birth year before 1920,whereas this reduction is moderate for households with birthyear between 1920 and 1945. It is also demonstrated thatdriving license holding choice is conditional on the carownership level choice, and that these two choices need to bemodeled in a dynamic framework.

The second part of the work investigates the performance ofthe mixed logit model using both simulated data and empiricalroute switching data. The empirical study mainly focused on theimpacts of information and incident related factors on drivers'route switching behavior.

The result shows that using mixed logit gives a significantimprovement in model performance as well as a more sensitiveexplanation of drivers' decision-making behavior. For apopulation with greatly varying tastes, simply using thestandard logit model to analyze its behavior can yield veryunrealistic results. However, care must be taken when settingthe number of random draws for simulating the choiceprobability of the mixed logit model in order to get reliableestimates.

The empirical results demonstrate that incident relatedfactors such as delay and information reliability havesignificant impacts on drivers' route switching, where themagnitude of the response to the change in the delay is shownto vary significantly between individuals. Other factors, suchas confidence in the estimated delay, gender, frequency of cardriving and attitude towards congestion, also make majorcontributions. In addition, it is found that individual's routeswitching behavior may differ depending on the purpose of thetrip and when the choice is made, i.e. pre-trip oren-route.

Keywords: car ownership, accessibility, logit model,route choice, heterogeneity, mixed logit model

nÃo hÃ
Sendo a intermediaÃÃo financeira a principal atividade dos bancos, alocando recursos de clientes superavitÃrios a clientes deficitÃrios, à na incerteza quanto ao carÃter e a capacidade de pagamento dos clientes que se estabelece o risco e com ele a necessidade de se buscar novas alternativas para se proteger de perdas potenciais, que podem refletir em menores lucros para as instituiÃÃes. AlÃm da subjetividade dos analistas de crÃdito, o uso de modelos quantitativos, baseados em prÃticas estatÃsticas, economÃtricas e matemÃticas, vÃm cada vez mais se firmando nos mercados como ferramenta de apoio aos gestores de crÃdito na tomada de decisÃo. VÃrios modelos de avaliaÃÃo de risco sÃo adotados pelas instituiÃÃes, modelos de credit scoring, behavioral scoring, sÃo exemplos destes modelos. O modelo de credit scoring tem sido um dos mais usados, em especial para concessÃo de crÃdito a pessoas fÃsicas. Os modelos de credit scoring utilizam tÃcnicas como a anÃlise de discriminantes, programaÃÃo matemÃtica, econometria, redes neurais, entre outras, para atravÃs da anÃlise de caracterÃsticas particulares dos indivÃduos, estabelecer uma mÃtrica de separaÃÃo de bons e maus pagadores, atribuindo probabilidades diferentes de inadimplÃncia aos mesmos. A presente dissertaÃÃo tem como objetivo central analisar os determinantes de inadimplÃncia (pessoa fÃsica), usando uma abordagem economÃtrica com base no modelo Logit. O modelo utilizado foi um modelo para aprovaÃÃo de crÃdito na abertura de conta corrente, partindo de um estudo com uma amostra de 308 observaÃÃes (cadastros pessoas fÃsicas), baseados na experiÃncia real de uma instituiÃÃo financeira, cujo objetivo à atingir uma taxa de aprovaÃÃo de crÃdito tal que a receita mÃdia depois das perdas de emprÃstimos seja maximizada.
In the financial intermediation, banks focus on its main activity, allocating resources from clients with surplus to deficit clients. The uncertainty related to the characteristics or payment capacity of the clients establishes the risk and the need to search for new alternatives to protect the institutions from potential losses, which may reflect on lower profits. Besides the subjective issue of credit analysts, the use of quantitative models, based on statistical, mathematical or econometric practices are becoming an important tool to support credit managers on the decision making process. There are several models of risk evaluation, which are adopted by financial institutions such as the credit scoring and the behavioral scoring models. The credit-scoring model has been widely used, especially on the concession of individual credit. The credit scoring model uses techniques such as discriminant analysis, mathematic programming, econometrics, neural networks, among others, to analyze particular characteristics of individuals where it establishes a metric separation of good and bad payers, therefore providing different nonpayment status to each. This present dissertation has the main objective of analyzing the determinants of nonpayment status (individuals), using an econometric approach based on the Logit model. The model utilized was a model for approval of credit in the opening from the bill shackle, starting from a study with 308 observations (physical registers Persons), based in the real experience of a financial institution, whose objective is he reach a credit approval rate such that the medium prescription after the losses of loans be maximized.

O presente trabalho procura avaliar a relevância da informação sobre a segurança dos automóveis para os consumidores brasileiros estimando um modelo de demanda logit aninhado controlado pelas notas de zero a cinco estrelas obtidas nos crash tests do Latin NCAP, pesquisa até o momento única no Brasil. Os resultados indicam que a nota nos crash tests são relevantes para a utilidade média dos consumidores com os carros mais bem avaliados mostrando um nível de utilidade maior em relação aos veículos menos seguros. Resultados auxiliares apontam que não há diferença nos preços médios dos modelos avaliados pelo Latin NCAP com notas de uma a cinco estrelas sendo os modelos avaliados com zero estrelas os únicos a apresentarem preços estatisticamente diferentes e mais baratos. Outros resultados obtidos mostram que um aumento no número de automóveis testados de uma determinada montadora aumenta mais que proporcionalmente a probabilidade de um outro automóvel da mesma marca ser avaliado
The present work aims to evaluate how important is the information about car safety for consumers in brazilian market estimating a nested logit demand model using the Latin NCAP\'s grades from zero to five stars as independent variables, yet this research is unique in Brazil. The results obtained show that Latin NCAP\'s grades show some relevance for the consumer\'s mean utiliy as the higher grades exibit higher mean utility levels related to the lowest grades. Auxiliary results show that there is no statistic diference among the car\'s prices evaluated with one to five stars in the crash tests as the automobiles evaluated with zero stars appear as being the cheapest of all. Others results show that an increase in the number of tested vehicles from a given manufacturer raises more than proportionally the probability of a car from this same brand to be tested

The aim of this study was to develop a framework to model travel choice behavior in order to estimate intercity travel demand at nation-level in the United States. Nested and mixed logit models were developed to study national-level intercity transportation in the United States. A separate General Aviation airport choice model to estimates General Aviation person-trips and number of aircraft operations though more than 3000 airports was also developed. The combination of the General Aviation model and the logit models gives the capability to estimate a full spectrum of intercity travel demand in the United States. The logit models were calibrated using a nationwide revealed preference survey (1995 American Travel Survey). Separate models were developed for business and non-business trip purposes. An airport choice model is integrated into the mode choice model to estimate both the market share between any origin-destination pair and other modes of transportation, and the market share split between airports associated with the origin-destination pairs. The explanatory variables used in the utility functions of the models are travel time, travel cost, and travelerâ s household income. The logit models are used to estimate the market share of automobile and commercial air transportation between 3091 counties and 443 commercial service airports in the United States. The model was also used to estimate market share for on-demand air taxi services. Given an input county-to-county trip demand table, the models were used to estimate county-to-county travel demand by automobile and commercial airline between all counties and commercial service airports in the United States. The model has been integrated into a computer software framework called the Transportation Systems Analysis Model (TSAM) that estimates nationwide intercity travel demand in the United States.
Ph. D.

Discrete choice models (DCMs) are used to forecast demand in a variety of engineering, marketing, and policy contexts, and understanding the uncertainty associated with model forecasts is crucial to inform decision-making. This thesis evaluates the suitability of DCMs for forecasting automotive demand. The entire scope of this investigation is too broad to be covered here, but I explore several elements with a focus on three themes: defining how to measure forecast accuracy, comparing model specifications and forecasting methods in terms of prediction accuracy, and comparing the implications of model specifications and forecasting methods on vehicle design. Specifically I address several questions regarding the accuracy and uncertainty of market share predictions resulting from choice of utility function and structural specification, estimation method, and data structure assumptions. I1 compare more than 9,000 models based on those used in peer-reviewed literature and academic and government studies. Firstly, I find that including more model covariates generally improves predictive accuracy, but that the form those covariates take in the utility function is less important. Secondly, better model fit correlates well with better predictive accuracy; however, the models I construct— representative of those in extant literature— exhibit substantial prediction error stemming largely from limited model fit due to unobserved attributes. Lastly, accuracy of predictions in existing markets is neither a necessary nor sufficient condition for use in design. Much of the econometrics literature on vehicle market modeling has presumed that biased coefficients make for bad models. For purely predictive purposes, the drawbacks of potentially mitigating bias using generalized method of moments estimation coupled with instrumental variables outweigh the expected benefits in the experiments conducted in this dissertation. The risk of specifying invalid instruments is high, and my results suggest that the instruments frequently used in the automotive demand literature are likely invalid. Furthermore, biased coefficients are not necessarily bad for maximizing the predictive power of the model. Bias can even aid predictions by implicitly capturing persistent unobserved effects in some circumstances. Including alternative specific constants (ASCs) in DCM utility functions improves model fit but not necessarily forecast accuracy. For frequentist estimated models all tested methods of forecasting ASCs improved share predictions of the whole midsize sedan market over excluding ASC in predictions, but only one method results in improved long term new vehicle, or entrant, forecasts. As seen in a synthetic data study, assuming an incorrect relationship between observed attributes and the ASC for forecasting risks making worse forecasts than would be made by a model that excludes ASCs entirely. Treating the ASCs as model parameters with full distributions of uncertainty via Bayesian estimation is more robust to selection of ASC forecasting method and less reliant on persistent market structures, however it comes at increased computational cost. Additionally, the best long term forecasts are made by the frequentist model that treats ASCs as calibration constants fit to the model post estimation of other parameters.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior
FundaÃÃo de Amparo à Pesquisa do Estado do CearÃ
How are choices made? The understanding of this matter is crucial, for instance, for an automobile industry firm that intends to launch a new model on the market, or for a political party that needs to decide which candidate will run for the office. It happens that buying the A or B car or voting for the A or candidate are usually individual choices. Therefore, when one understands how the choices are made, the chances of selling more cars or winning the election are increased greatly. That is the primary subject of this study: choices. More specifically, it deals with choices at the light of the Economical Theory. And in order to lead sequence of three empiric exercises about the theme a forefront technique named Mixed LOGIT is used. In the first exercise the set of the peopleâs choices contains options for the buying of automobiles, and how the choices are made in that consumption environment will be studied. Consequently, it will be possible to compute ownprice- elasticity, cross-price-elasticity, markups and surplus of the consumer and the producer. In the second exercise the set of choices will represent options of kinds workers, and how the firms make the choices on how to fill the job vacancies will be studied. Consequently, it will be possible to understand how the social networks can affect the labor market. In the third exercise the set of choices will represent the Presidential candidates running for the 2006 election, and how the voters can be induced to voting to a certain candidate will be studied. Consequently, it will be possible to estimate how a cash transfer programme, such as Programa Bolsa FamÃlia (Family Grant Programme) can influence the result of the election.
Como as pessoas fazem suas escolhas? O entendimento desta questÃo à crucial, por exemplo, para uma montadora de veÃculos que pretende lanÃar um novo modelo de automÃvel no mercado, ou para um partido polÃtico que precisa decidir qual candidato lanÃar. Pois comprar um carro A ou B ou votar num candidato A ou B costumam ser decisÃes de escolha individual, entÃo, ao se entender como as escolhas sÃo feitas se aumentam as chances de sucesso de venda ou de votaÃÃo. Esta tese trata fundamentalmente disso: escolhas. Mais especificamente, trata de escolhas à luz da Teoria EconÃmica. E para conduzir uma seqÃÃncia de trÃs exercÃcios empÃricos sobre o tema usa-se uma tÃcnica de vanguarda denominada de Mixed LOGIT. No primeiro exercÃcio o conjunto de escolhas das pessoas contÃm opÃÃes para compra de automÃveis, e se estudarà como sÃo feitas as escolhas nesse ambiente de consumo. Por conseqÃÃncia desse entendimento, serà possÃvel computar elasticidades-preÃo, preÃo-cruzado, markups e excedentes do consumidor e do produtor. No segundo exercÃcio o conjunto de escolhas representarà opÃÃes de tipos de trabalhadores, e se estudarà como as firmas fazem escolhas para ocupar postos de trabalho vagos. Por conseqÃÃncia desse entendimento, serà possÃvel compreender como as redes sociais de trabalhadores podem afetar o mercado de trabalho. E no terceiro exercÃcio o conjunto de escolhas representarà os candidatos a PresidÃncia da RepÃblica nas eleiÃÃes de 2006, e se estudarà como os eleitores podem ser induzidos a votar em determinado candidato. Por conseqÃÃncia desse entendimento, serà possÃvel estimar o quanto um programa de transferÃncia direta de renda como o Bolsa FamÃlia à capaz de influenciar o resultado de uma eleiÃÃo.

Submitted by Mônica Correia Aquino (monicacorreiaaquino@gmail.com) on 2011-07-04T20:33:51Z No. of bitstreams: 1 2009_tese_fcpLourenço.txt: 299561 bytes, checksum: c7977ba607a6ce65e8b619dc710f5da5 (MD5)
Approved for entry into archive by Mônica Correia Aquino(monicacorreiaaquino@gmail.com) on 2011-07-04T20:34:14Z (GMT) No. of bitstreams: 1 2009_tese_fcpLourenço.txt: 299561 bytes, checksum: c7977ba607a6ce65e8b619dc710f5da5 (MD5)
Made available in DSpace on 2011-07-04T20:34:14Z (GMT). No. of bitstreams: 1 2009_tese_fcpLourenço.txt: 299561 bytes, checksum: c7977ba607a6ce65e8b619dc710f5da5 (MD5) Previous issue date: 2009
How are choices made? The understanding of this matter is crucial, for instance, for an automobile industry firm that intends to launch a new model on the market, or for a political party that needs to decide which candidate will run for the office. It happens that buying the A or B car or voting for the A or B candidate are usually individual choices. Therefore, when one understands how the choices are made, the chances of selling more cars or winning the election are increased greatly. That is the primary subject of this study: choices. More specifically, it deals with choices at the light of the Economical Theory. And in order to lead a sequence of three empiric exercises about the theme a forefront technique named Mixed LOGIT is used. In the first exercise the set of the people’s choices contains options for the buying of automobiles, and how the choices are made in that consumption environment will be studied. Consequently, it will be possible to compute ownprice- elasticity, cross-price-elasticity, markups and surplus of the consumer and the producer. In the second exercise the set of choices will represent options of kinds of workers, and how the firms make the choices on how to fill the job vacancies will be studied. Consequently, it will be possible to understand how the social networks can affect the labor market. In the third exercise the set of choices will represent the Presidential candidates running for the 2006 election, and how the voters can be induced to voting to a certain candidate will be studied. Consequently, it will be possible to estimate how a cash transfer programme, such as Programa Bolsa Família (Family Grant Programme) can influence the result of the election
Como as pessoas fazem suas escolhas? O entendimento desta questão é crucial, por exemplo, para uma montadora de veículos que pretende lançar um novo modelo de automóvel no mercado, ou para um partido político que precisa decidir qual candidato lançar. Pois comprar um carro A ou B ou votar num candidato A ou B costumam ser decisões de escolha individual, então, ao se entender como as escolhas são feitas se aumentam as chances de sucesso de venda ou de votação. Esta tese trata fundamentalmente disso: escolhas. Mais especificamente, trata de escolhas à luz da Teoria Econômica. E para conduzir uma seqüência de três exercícios empíricos sobre o tema usa-se uma técnica de vanguarda denominada de Mixed LOGIT. No primeiro exercício o conjunto de escolhas das pessoas contém opções para compra de automóveis, e se estudará como são feitas as escolhas nesse ambiente de consumo. Por conseqüência desse entendimento, será possível computar elasticidades-preço, preço-cruzado, markups e excedentes do consumidor e do produtor. No segundo exercício o conjunto de escolhas representará opções de tipos de trabalhadores, e se estudará como as firmas fazem escolhas para ocupar postos de trabalho vagos. Por conseqüência desse entendimento, será possível compreender como as redes sociais de trabalhadores podem afetar o mercado de trabalho. E no terceiro exercício o conjunto de escolhas representará os candidatos a Presidência da República nas eleições de 2006, e se estudará como os eleitores podem ser induzidos a votar em determinado candidato. Por conseqüência desse entendimento, será possível estimar o quanto um programa de transferência direta de renda como o Bolsa Família é capaz de influenciar o resultado de uma eleição.

Submitted by Kleber Osman (kleber.osman@icloud.com) on 2018-03-02T01:11:37Z No. of bitstreams: 1 Dissertação - Kleber Osman.pdf: 857853 bytes, checksum: 51c38d1dfdb2e92c87e2269bcc638896 (MD5)
Rejected by Thais Oliveira (thais.oliveira@fgv.br), reason: Prezado Kleber, boa noite! Para que possamos aprovar seu trabalho, será necessário que faça somente uma alteração: - Na capa, o nome "GETULIO" não tem acento. Por gentileza alterar e submeter novamente. Obrigada! on 2018-03-06T00:00:13Z (GMT)
Submitted by Kleber Osman (kleber.osman@icloud.com) on 2018-03-06T02:15:42Z No. of bitstreams: 1 Dissertação - Kleber Osman.pdf: 861905 bytes, checksum: 3478b1be8a0bd0efd15781f0c21be7cd (MD5)
Approved for entry into archive by Thais Oliveira (thais.oliveira@fgv.br) on 2018-03-06T19:29:17Z (GMT) No. of bitstreams: 1 Dissertação - Kleber Osman.pdf: 861905 bytes, checksum: 3478b1be8a0bd0efd15781f0c21be7cd (MD5)
Made available in DSpace on 2018-03-07T13:05:15Z (GMT). No. of bitstreams: 1 Dissertação - Kleber Osman.pdf: 861905 bytes, checksum: 3478b1be8a0bd0efd15781f0c21be7cd (MD5) Previous issue date: 2018-02-06
A crise previdenciária vivenciada no decorrer dos últimos anos mostrou que seus impactos não se restringem ao desequilíbrio das contas públicas devido ao aumento dessas despesas obrigatórias, mas também efeitos sociais e econômicos negativos podem atingir toda a economia brasileira. Diante disso, esforços se direcionaram na tentativa de encontrar uma alternativa que permita desonerar as contas públicas para evitar uma crise na economia local, assim como garantir que os aposentados mantenham a mesma renda frente ao salário recebido anteriormente à aposentadoria. Neste sentido, este estudo investiga o perfil dos investidores em previdência privada, visando detectar as principais características que os diferenciam. A análise do perfil dos indivíduos que contribuem para planos de previdência privada permite que políticas de incentivo a este produto sejam melhores desenhadas e, possibilita à iniciativa privada o seu direcionamento ao público-alvo de forma mais eficiente – uma vez reveladas as variáveis socioeconômicas de maior relevância. O método utilizado na indicação desses fatores baseia-se na estimação LOGIT, que tem como objetivo principal auferir quais são as variáveis com maiores ganhos marginais que afetam a probabilidade do indivíduo participar de plano de previdência privada. E, com base nos resultados, constatou-se que as variáveis 'educação' e 'renda' são preponderantes para a inscrição do indivíduo em plano de previdência privada.
The social security crisis experienced in recent years has shown that its impacts are not restricted to the imbalance of public accounts due to the increase in these compulsory expenses, but also negative social and economic effects can affect the entire Brazilian economy. In the face of this, efforts were directed towards finding an alternative that would allow public accounts to be discharged to avoid a crisis in the local economy, as well as ensuring that retirees maintain the same income against the salary received prior to retirement. In this sense, this study investigates the profile of private pension investors in order to detect the main characteristics that differentiate them. The analysis of the profile of individuals contributing to private pension plans allows policies to encourage this product to be better designed and allows the private sector to target the target public more efficiently - once the socioeconomic variables of greater relevance. The method used to indicate these factors is based on the LOGIT estimation, whose main objective is to determine which are the variables with the greatest marginal gains that affect the probability of the individual participating in a private pension plan. And, based on the results, it was verified that the variables 'education' and 'income' are preponderant for the inscription of the individual in private pension plan.

La crisi finanziaria del 2007-2009 non solo ha attaccato l’economia globale in maniera diversificata, ma ha anche generato effetti differenti per i diversi paesi colpiti. Proprio avendo preso atto della crisi di liquidità, che si è profilata in Europa e non solo, e della presenza di situazioni di estrema difficoltà in cui versano i mercati finanziari, si è dovuto riscoprire una tipologia di rischio che prima della crisi era stato considerato secondariamente, si tratta del rischio sistemico. Questa tipologia di rischio non scaturisce, come si pensava inizialmente, solo a seguito di crisi di tipo valutario o di crisi bancarie, ma oltre alle banche ha colpito anche altri settori come hedge funds, assicurazioni, brokers ecc. Per studiare la salute del nostro sistema economico bisogna andare ad analizzare le relazioni esistenti all’interno di un sistema in cui autorità di vigilanza, intermediari finanziari e destinatari dei servizi offerti sono sempre più legati tra loro. Il legame tra queste figure che lavorano nei mercati è così forte che uno shock, che può essere creato da una qualsiasi tipologia di evento negativo, riesce a diffondersi talmente velocemente da riuscire a produrre quell’effetto definito “effetto domino” che va a colpire tutti gli intermediari presenti sul mercato. A seguito di questa “reazione a catena” le tensioni che si vengono a generare a causa del fallimento di una società di intermediazione finanziaria o di una banca, si ripercuotono non solo sugli altri intermediari ma anche nei diversi mercati, provocando il fallimento di tutti quei soggetti che sono presenti sul mercato. La crisi è riuscita a mettere in risalto come gli operatori di mercato non siano capaci di reggere un sistema che, grazie ai progressi tecnologici dell’informazione e ai diffusissimi mezzi di comunicazione, risulta effettivamente complesso. Per questo motivo organizzazioni internazionali e le banche centrali hanno concentrato sforzi e tempo per incrementare modelli e strumenti che siano d’aiuto al fine di valutare, individuare e monitorare costantemente possibili minacce che possono andare ad intaccare la stabilità del sistema. La recente crisi finanziaria ha messo in risalto quelli che sono i limiti del nostro sistema, sia dal punto di vista della regolamentazione, che da quello della supervisione degli intermediari e dei mercati. Il problema più evidente è dato dal fatto che si è sempre posta l’attenzione sulla solidità delle singole componenti che compongono il sistema finanziario, cioè inizialmente si è sempre avuto un approccio micro-prudenziale anziché tenere presente la stabilità del sistema nella sua complessità. La crisi infatti ha mostrato come il sistema è stato vulnerabile, tant’è vero che shock che si ritenevano circoscritti a determinati settori, hanno infine intaccato l’intero sistema.

Master of Science
Department of Statistics
Perla E. Reyes Cuellar
Data with ordinal responses may be encountered in many research fields, such as social, medical, agriculture or financial sciences. In this paper, we present a case study on cumulative logit models with low frequency and mixed effects and discuss some strengths and limitations of the current methodology. Two plant pathologists requested our statistical advice to fit a cumulative logit mixed model seeking for the effect of six commercial products on the control of a seed and seedling disease in soybeans in vitro. In their attempt to estimate the model parameters using a generalized linear mixed model approach with PROC GLIMMIX, the model failed to converge. Three alternative approaches to solve the problem were examined: 1) stratifying the data searching for the random effect; 2) assuming the random effect would be small and reducing the model to a fixed model; and 3) combining the original categories of the response variable to a lower number of categories. In addition, we conducted a power analysis to evaluate the required sample size to detect treatment differences. The results of all the proposed solutions were similar. Collapsing categories for a cumulative/proportional odds model has little effect on estimation. The sample size used in the case study is enough to detect a large shift of frequencies between categories, but not for moderated changes. Moreover, we do not have enough information to estimate a random effect. Even when it is present, the results regarding the fixed factors: pathogen, evaluation day, and treatment effects are the same as the obtained by the fixed model alternatives. All six products had a significant effect in slowing the effect of the pathogen, but the effects vary between pathogen species and assessment timing or date.

MENDONÇA, Kamila Vieira de. Análise das causas socioeconômicas da pobreza rural no Ceará. 2009. 69 f.: Dissertação (mestrado) - Universidade Federal do Ceará. Mestrado Acadêmico em Economia Rural (MAER). Centro de Ciências Agrárias. Fortaleza-CE. 2009
Submitted by Francisco Helder Macêdo Rangel (fhelder@ufc.br) on 2014-09-25T12:35:25Z No. of bitstreams: 1 2009_dis_kvmendonça.pdf: 414467 bytes, checksum: e7ff04133b8341eea985a7eb72592dc8 (MD5)
Approved for entry into archive by Margareth Mesquita(margaret@ufc.br) on 2014-10-02T13:55:35Z (GMT) No. of bitstreams: 1 2009_dis_kvmendonça.pdf: 414467 bytes, checksum: e7ff04133b8341eea985a7eb72592dc8 (MD5)
Made available in DSpace on 2014-10-02T13:55:35Z (GMT). No. of bitstreams: 1 2009_dis_kvmendonça.pdf: 414467 bytes, checksum: e7ff04133b8341eea985a7eb72592dc8 (MD5) Previous issue date: 2009
The poverty aims to analyse in the rural areas in the State of the Ceará in which it refers to his causes socioeconomics, like education, age, color, type, principal activity of occupation and government transfers. Besides, there aim to analyse the characteristics of the rural residences made a list to the acess the public goods. There are studied the causes of the rural pverty of Ceará with the objective to check possible alterations in the level of the poverty in the previous and subsequent period of the Real Plan. From the microdata of the PNAD, obtained near the IBGE, there are determined the variables that present association with the rural poverty, through the model of regression logit and to analyse the determinants of the individuals, the marginal contributions are calculated. Besides, the public goods are calculated the descriptive statistics of the variable made a list to the causes of the poverty and to the acess. It is ended that the variables sex, level of schooling, receiving of retirement, boarding house and principal activity of work have significant relation with the poverty o the region, being the principals to education and the government transfers. Besides, it increased the acess to the public goods, this fact that provided better quality of life for the rural population of Ceará.
Objetiva-se analisar a pobreza nas áreas rurais no Estado do Ceará no que se refere às suas causas socioeconômicas como educação, idade, cor gênero, atividade principal de ocupação e transferências governamentais. Além disso, objetiva-se analisar as características dos domicílios rurais relacionadas ao acesso a bens públicos. Estudam-se as causas da pobreza rural cearense com o objetivo de verificar possíveis alterações no nível da pobreza no período anterior e posterior do Plano Real. A partir dos microdados das PNAD, obtidos junto ao IBGE, determinam-se as variáveis que apresentam associações com a pobreza rural, por meio do modelo de regressão Logit e para analisar os determinantes dos indivíduos, calculam-se as contribuições marginais. Ademais, são calculados as estatísticas descritivas das variáveis relacionadas às causas da pobreza e ao acesso a bens públicos. Conclui-se que as variáveis sexo, nível de escolaridade, recebimento de aposentadoria, pensão e atividade principal de trabalho possuem relação significativa com a pobreza da região, sendo as principais a educação e as transferências governamentais. Além disso, aumentou o acesso aos bens públicos, fato este que proporcionou melhor qualidade de vida para a população rural cearense.

The importance of traffic planning has, throughout the years, been in- creased, providing sustainable developments of traffic and infrastructural investments. The analysis of the current traffic situation and the evalua- tion of the effects of a future investment are crucial for the socio-economic benefits maintenance. These analyses and evaluations are most commonly done using traffic simulation models. One of the main traffic planning aims, nowadays, is to increase the number of public transport users against the number of private car users. This change in mode choice is called transition effect and could be beneficial both from an environmental and socio-economic perspective. This thesis aims to evaluate and improve the macroscopic traffic demand and transition model, used fot the city of Norrköping. Additionally, the thesis investigates if a general transition Logit model can be developed and which parameters are the most important to be included in a modal choice estimation. For the needs of this study, the traffic planning software Visum is used. The travel mode distribution is calculated by Logit models coded in Python-scripts integrated in Visum. Then, a traffic assignment is performed by Visum, computing new travel times as inputs to the Logit model and this iterative procedure continues until the system reaches an equilibrium. The thesis aims for a more reliable prediction of the transition effect by correcting the Python-scripts and estimating the parameters of the Logit model using data from surveys. The study shows that travel times is the most important factor for realistic results generation. However, the data used for the estimation of the Logit model parameters did not include sufficient information of travel times. The travel times had to be calculated, using two different methods, in order to be included in the estimation of new parameters. Although these methods could not provide any positive effects on the transition, they did prove the importance and significance travel time have when developing a traffic model. The result of the study invokes the importance to further develop the method of calculating travel times, when the input data is not sufficient, and shows that the travel time parameters are case specific.

Ingeniero Civil
Los modelos de elección discreta basados en utilidades o disposiciones a pagar han sido muy utilizados para representar el comportamiento de los consumidores en sistemas de Uso de Suelo y de Transporte. En este trabajo se desarrolla un modelo de elección que usa disposiciones a pagar para representar la elección en mercados de remate, donde los bienes se asignan a los mejores postores. Los modelos tradicionales no consideran el hecho de que estas funciones están siempre acotadas superiormente porque los consumidores tienen restricciones presupuestarias. Si bien se han generado enfoques para representar esta situación, estas son heurísticas solamente. En este trabajo se deduce de un modo riguroso un nuevo modelo de elección discreta binomial donde las funciones de valor son disposiciones a pagar acotadas superiormente. En particular, se obtiene una expresión analítica en el caso en que estas funciones son variables aleatorias idénticamente distribuidas Gumbel. Este modelo junto con incorporar en su formulación la influencia de la restricción presupuestaria, tiene la interesante propiedad de permitir estimar un número mayor de parámetros que en el caso del modelo Logit Binomial clásico. Junto con caracterizar este modelo, se probó empíricamente usando bases sintéticas que es posible estimar sus parámetros y que en muchos casos, entrega estimaciones más precisas que el modelo Logit Binomial tradicional.

Made available in DSpace on 2017-07-10T18:33:41Z (GMT). No. of bitstreams: 1 Carlos A Goncalves Junior.pdf: 339511 bytes, checksum: 6c64752683f2e8ade22f7ad3f541ffd8 (MD5) Previous issue date: 2009-04-16
This research work aims to create a tool that can help a credit cooperative in the region of Toledo in the analysis and concession of agricultural credit calculating the probability of completion of the contracts, what permits to predict a possible default, using the logistic regression model - Logit. The theoretical framework used, based on the Theory of Transaction Costs, identifies the default as the result of the incompleteness of contracts and of the asymmetry of information between the parties in a transaction. The goal is that this tool can reduce the asymmetry of information between the borrowers and the credit cooperative, in order to avoid the granting of credit to possible defaulters. To do so, it was collected information about the borrowers from the records of the cooperative from the year 2004 to 2007, aiming to draw a profile of the credit borrower. Later, it was estimated the logistic regression model for 10 different samples to identify the sample that received the greatest number of his between payers and defaulters. It was established that the estimated model was more efficient to identify the payers contracts than defaulters contracts, and even with a not too high percentage average of correct the model can assist the decision-making of the cooperative in granting credit.
Este trabalho de pesquisa objetiva criar uma ferramenta que possa auxiliar uma cooperativa de crédito da região de Toledo na análise e concessão de crédito agrícola calculando a probabilidade de cumprimento dos contratos, o que permite antever os possíveis contratos inadimplentes, utilizando-se o modelo de regressão logística Logit. O referencial teórico utilizado, baseado na Teoria dos Custos de Transação, identifica a inadimplência como sendo resultado da incompletude dos contratos e da assimetria de informações entre as partes envolvidas em uma transação. O objetivo é que essa ferramenta possa reduzir a assimetria de informações entre os tomadores e a cooperativa de crédito, no intuito de evitar a concessão de crédito a possíveis inadimplentes. Para isso, coletou-se junto à cooperativa informações cadastrais dos tomadores, no período de 2004 a 2007, objetivando traçar um perfil do tomador de crédito. Posteriormente estimou-se o modelo de regressão logística para 10 amostras aleatórias diferentes, a fim de identificar a amostra que obtivesse o maior número de acertos entre adimplentes e inadimplentes. Constatou-se que o modelo estimado foi mais eficiente para identificar os contratos adimplentes que os inadimplentes, e que mesmo com um porcentual médio de acerto não muito elevado o modelo pode auxiliar a tomada de decisão da cooperativa na concessão de crédito.