Página 1 dos resultados de 17 itens digitais encontrados em 0.050 segundos

‣ Modelos de séries temporais aplicados à análise prospectiva de concessão de crédito bancário; Time series models applied to forecast analysis of banking credit concessions

Abitante, Kleber Giovelli
Fonte: Biblioteca Digitais de Teses e Dissertações da USP Publicador: Biblioteca Digitais de Teses e Dissertações da USP
Tipo: Dissertação de Mestrado Formato: application/pdf
Publicado em 19/03/2007 Português
Relevância na Pesquisa
69.25247%
O presente trabalho teve por objetivo modelar as séries de concessão de crédito bancário às pessoas físicas, às pessoas jurídicas e para financiamento de atividades rurais, bem como realizar previsões a cerca dos comportamentos destas séries. A metodologia utilizada foi de Auto- Regressão Vetorial. A propriedade de co-integração entre as variáveis foi considerada no trabalho, sendo que foram estimados modelos de Auto-Regressão Vetorial com Correção de Erro – VEC. Os resultados mostram que o produto, a taxa de juros cobrada nos empréstimos, as exportações e as vendas no varejo podem auxiliar na geração de previsões satisfatórias das concessões de crédito às pessoas jurídicas e às pessoas físicas. Para o modelo de previsão das concessões de crédito para financiamento de atividades rurais, utilizaram-se variáveis referentes à produção de fertilizantes, vendas de tratores e colheitadeiras, produção de leite e produção de carnes bovina, suínas e de aves, sendo que as previsões geradas pelo modelo apresentaram performance adequada, dada a dificuldade da modelagem.; The aim of this study was to model the series of banking credit concessions to individuals, to firms and for rural activities financing...

‣ Avaliação da previsão hidroclimática no Alto Paraguai

Allasia Piccilli, Daniel Gustavo
Fonte: Universidade Federal do Rio Grande do Sul Publicador: Universidade Federal do Rio Grande do Sul
Tipo: Tese de Doutorado Formato: application/pdf
Português
Relevância na Pesquisa
19.094614%
Neste estudo foi analisada a previsibilidade hidroclimática na bacia do Alto Paraguai no curto e longo prazo. Nestas últimas décadas, a região foi marcada por uma forte variabilidade climática, passando por um período extremamente seco durante a década de 1960 e por um período extremamente úmido a partir do início da década de 1970. Desta forma, avaliar a previsibilidade hidroclimática é muito importante para o gerenciamento dos recursos hídricos na bacia, e, conseqüentemente do ecossistema que depende deste recurso. A previsibilidade de longo prazo da Bacia do Alto Paraguai foi analisada mediante técnicas estatísticas dado que a região do Brasil Central é considerada de baixa previsibilidade de longo prazo mediante modelos dinâmicos. O objetivo da análise foi o de explorar possíveis relações entre os fenômenos climáticos globais (representados pelos índices climáticos) sobre o clima da BAP com valor prognóstico. Os resultados da análise mostraram que o fenômeno que tem a maior capacidade de modulação do clima na Bacia do Alto Paraguai é a Oscilação Decadal do Pacífico (PDO). A PDO se caracteriza por fases de aproximadamente 25 anos nas quais o índice que a representa se encontra acima ou abaixo do valor normal. A duração das fases do PDO é definida por alterações climáticas abruptas que se encontraram muito bem representadas no Pantanal...

‣ Modelling and forecasting brent prices

Silva, Tânia Cristina Dinis Marques e
Fonte: Instituto Universitário de Lisboa Publicador: Instituto Universitário de Lisboa
Tipo: Dissertação de Mestrado
Publicado em //2010 Português
Relevância na Pesquisa
28.438516%
Mestrado em Finanças/ C23, Q43; Desde o início dos tempos que o Brent, mais conhecido por petróleo, tem sido uti-lizado em diversas aplicações, devido à sua elevada densidade energética, facilidade de transporte e relativa abundância. Nos últimos anos, o Brent tornou-se na fonte de energia mais importante, desempenhando um papel preponderante na manutenção da nossa actual sociedade. Neste contexto, o objectivo principal deste trabalho é mode-lar e prever os preços mensais e diários do Brent, de forma a melhor compreender e antever o seu comportamento. Na modelação e previsão dos preços utilizaram-se duas abordagens diferentes. A primeira baseia-se na análise de séries temporais com memória longa. A presença de memória longa é veri cada na média condicional e modelada a partir de modelos ARFIMA. Esta característica é também analisada na volatilidade da série e mode-lada através de modelos FIGARCH, FIAPARCH ou FIEGARCH. A outra abordagem considera modelos estocásticos de mudança de regime, nomeadamente modelos STAR, SETAR e MS-AR. A modelação dos preços diários de Brent é feita com base em modelos de séries tem-porais considerando memória longa, uma vez que esta característica foi identi cada na volatilidade da série. Modelos de mudança de regime foram também aplicados...

‣ The Association of Weather Variability and Under Five Malaria Mortality in KEMRI/CDC HDSS in Western Kenya 2003 to 2008: A Time Series Analysis

Sewe, Maquins; Rocklöv, Joacim; Williamson, John; Hamel, Mary; Nyaguara, Amek; Odhiambo, Frank; Laserson, Kayla
Fonte: MDPI Publicador: MDPI
Tipo: Artigo de Revista Científica
Português
Relevância na Pesquisa
37.760723%
Malaria is among the leading causes of mortality in the younger under-five group of children zero to four years of age. This study aims at describing the relationship between rainfall and temperature on under-five malaria or anaemia mortality in Kenya Medical Research Institute and United States Centers for Disease Control (KEMRI/CDC) Health and Demographic Surveillance System (HDSS). This study was conducted through the ongoing KEMRI and CDC collaboration. A general additive model with a Poisson link function was fit to model the weekly association of lagged cumulative rainfall and average temperature on malaria/anemia mortality in KEMRI/CDC HDSS for the period 2003 to 2008. A trend function was included in the model to control for time trends and seasonality not explained by weather fluctuations. 95% confidence intervals was presented with estimates. Malaria or anemia mortality was found to be associated with changes in temperature and rainfall in the KEMRI HDSS, with a delay up to 16 weeks. The empirical estimates of associations describe established biological relationships well. This information, and particularly, the strength of the relationships over longer lead times can highlight the possibility of developing a predictive forecast with lead times up to 16 weeks in order to enhance preparedness to high transmission episodes.

‣ GIStorage: um serviço de informação para grades com suporte a algoritmos de predição de desempenho

Orengo, Jean Paulo Sandri
Fonte: Pontifícia Universidade Católica do Rio Grande do Sul; Porto Alegre Publicador: Pontifícia Universidade Católica do Rio Grande do Sul; Porto Alegre
Tipo: Dissertação de Mestrado
Português
Relevância na Pesquisa
27.76982%
Para alocar recursos e submeter tarefas numa grade computacional serviços de descoberta, alocação e escalonamento precisam conhecer o desempenho dos recursos. Como as tarefas serão executadas num momento futuro, estes serviços podem empregar algoritmos de predição para prever o desempenho dos recursos, melhorando a qualidade de suas decisões. Além disso, algoritmos de predição baseados em séries temporais demandam informações históricas sobre o desempenho dos recursos para prever o comportamento futuro dos mesmos. Para dar suporte a algoritmos e serviços de predição é proposto o GIStorage, um serviço de informação para grades computacionais projetado para armazenar informações sobre recursos. O GIStorage é baseado no modelo GMA, sendo estruturado como uma árvore para obter bom desempenho e armazenar grande volume de dados.; In order to allocate resources and submit jobs to a grid, resource discovery and scheduling services need to know in advance the performance of the candidate resources. Since jobs will be executed in a future time, these services may use prediction algorithms to forecast resources performance, improving the quality of their decisions. Additionally, prediction algorithms that use time series analysis demand historical performance information to predict future behavior. To support prediction algorithms and services we propose GIStorage...

‣ Dowscaling estocástico para extremos climáticos via interpolação espacial

Carvalho, Daniel Matos de
Fonte: Universidade Federal do Rio Grande do Norte; BR; UFRN; Programa de Pós-Graduação em Matemática Aplicada e Estatística; Probabilidade e Estatística; Modelagem Matemática Publicador: Universidade Federal do Rio Grande do Norte; BR; UFRN; Programa de Pós-Graduação em Matemática Aplicada e Estatística; Probabilidade e Estatística; Modelagem Matemática
Tipo: Dissertação Formato: application/pdf
Português
Relevância na Pesquisa
18.614114%
Present day weather forecast models usually cannot provide realistic descriptions of local and particulary extreme weather conditions. However, for lead times of about a small number of days, they provide reliable forecast of the atmospheric circulation that encompasses the subscale processes leading to extremes. Hence, forecasts of extreme events can only be achieved through a combination of dynamical and statistical analysis methods, where a stable and significant statistical model based on prior physical reasoning establishes posterior statistical-dynamical model between the local extremes and the large scale circulation. Here we present the development and application of such a statistical model calibration on the besis of extreme value theory, in order to derive probabilistic forecast for extreme local temperature. The dowscaling applies to NCEP/NCAR re-analysis, in order to derive estimates of daily temperature at Brazilian northeastern region weather stations; Conselho Nacional de Desenvolvimento Científico e Tecnológico; Os dados de reanálise de temperatura do ar e precipitação do NCEP National Centers for Environmental Predictions serão refinados para a produção dos níveis de retorno para eventos extremos nas 8 capitais do Nordeste Brasileiro - NB: São Luis...

‣ A statistical-dynamical approach to intraseasonal prediction of tropical cyclogenesis in the western North Pacific

Mundhenk, Bryan D.
Fonte: Monterey, California. Naval Postgraduate School Publicador: Monterey, California. Naval Postgraduate School
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
17.94459%
Approved for public release, distribution unlimited; We have developed a combined statistical-dynamical prediction scheme to predict the probability of tropical cyclone (TC) formation at daily, 2.5° horizontal resolution across the western North Pacific at intraseasonal lead times. Through examination of previous research and our own analysis, we chose five variables to represent the favorability of the climate system to support tropical cyclogenesis. These so-called large-scale environmental factors (LSEFs) include: low-level relative vorticity, sea surface temperature, vertical wind shear, Coriolis, and upper-level divergence. Logistic regression was employed to generate a statistical model representing the probability of TC formation at every grid point based on these LSEFs. Thorough verification of zero-lead hindcasts reveals this model displays skill and potential value for risk adverse customers. In particular, these hindcasts had a positive Brier skill score of 0.03 and a skillful relative operating characteristic skill score of 0.68. The fully coupled, one-tier NCEP Climate Forecast System was used as the dynamical model with which to forecast the LSEFs and, in turn, force the regression model. A series of individual TC case studies were conducted to demonstrate the predictive potential...

‣ A multivariate times series analysis of U.S. Army recruiting

Burger, Eric C.
Fonte: Monterey, California. Naval Postgraduate School Publicador: Monterey, California. Naval Postgraduate School
Tipo: Tese de Doutorado
Português
Relevância na Pesquisa
58.009106%
The United States Army Recruiting Command requires tools to quantify the impact of factors in the recruiting environment, identify differences in the recruiting processes across its five regional subordinate units, and measure the effectiveness of its policies and resource expenditures. This thesis examines recruiting data for the "high-quality" male demographic from July 1992 to September 1997. It uses multivariate time series analysis to predict the number of enlistment contracts signed in a month as a function of fifteen exogenous and endogenous factors plus monthly indicators. A stepwise recursion using bootstrap simulation is developed to identify significant factors in the multivariate time series. The significant factors in the reduced models are compared to those contained in models developed in previous studies. The models are also used to create nine- month projections of recruiting production, which are compared to known production figures from test set data to determine forecast accuracy. The results of this research support the intuition that the influential factors differ by region. The stepwise model reduction recursion using bootstrap simulation offers potential for further refinement and application.

‣ Bayesian Methods for Completing Data in Spatial Models

Llano Verduras, Carlos; Polasek, Wolfgang; Sellner, Richard
Fonte: Rimini Centre for Economic Analysis Publicador: Rimini Centre for Economic Analysis
Tipo: Artigo de Revista Científica
Português
Relevância na Pesquisa
37.89268%
Completing data sets that are collected in heterogeneous units is a quite frequent problem. Chow and Lin (1971) were the first to develop a unified framework for the three problems (interpolation, extrapolation and distribution) of predicting times series by related series (the ‘indicators’). This paper develops a spatial Chow-Lin procedure for cross-sectional data and compares the classical and Bayesian estimation methods. We outline the error covariance structure in a spatial context and derive the BLUE for ML and Bayesian MCMC estimation. In an example, we apply the procedure to Spanish regional GDP data between 2000 and 2004. We assume that only NUTS-2 GDP is known and predict GDP at NUTS-3 level by using socio-economic and spatial information available at NUTS-3. The spatial neighborhood is defined by either km distance, travel time, contiguity or trade relationships. After running some sensitivity analysis, we present the forecast accuracy criteria comparing the predicted values with the observed ones.

‣ Essays in energy economics: The electricity industry

Martinez-Chombo, Eduardo
Fonte: Universidade Rice Publicador: Universidade Rice
Tipo: Thesis; Text Formato: 161 p.; application/pdf
Português
Relevância na Pesquisa
17.94459%
Electricity demand analysis using cointegration and error-correction models with time varying parameters: The Mexican case. In this essay we show how some flexibility can be allowed in modeling the parameters of the electricity demand function by employing the time varying coefficient (TVC) cointegrating model developed by Park and Hahn (1999). With the income elasticity of electricity demand modeled as a TVC, we perform tests to examine the adequacy of the proposed model against the cointegrating regression with fixed coefficients, as well as against the spuriousness of the regression with TVC. The results reject the specification of the model with fixed coefficients and favor the proposed model. We also show how some flexibility is gained in the specification of the error correction model based on the proposed TVC cointegrating model, by including more lags of the error correction term as predetermined variables. Finally, we present the results of some out-of-sample forecast comparison among competing models. Electricity demand and supply in Mexico. In this essay we present a simplified model of the Mexican electricity transmission network. We use the model to approximate the marginal cost of supplying electricity to consumers in different locations and at different times of the year. We examine how costs and system operations will be affected by proposed investments in generation and transmission capacity given a forecast of growth in regional electricity demands. Decomposing electricity prices with jumps. In this essay we propose a model that decomposes electricity prices into two independent stochastic processes: one that represents the "normal" pattern of electricity prices and the other that captures temporary shocks...

‣ Bayesian analysis of autoregressive panel data model: application in genetic evaluation of beef cattle; Análise Bayesiana do modelo auto-regressivo para dados em painel: aplicação na avaliação genética de bovinos de corte

SILVA, Fabyano Fonseca e; SÁFADI, Thelma; MUNIZ, Joel Augusto; ROSA, Guilherme Jordão Magalhães; AQUINO, Luiz Henrique de; MOURÃO, Gerson Barreto; SILVA, Carlos Henrique Osório
Fonte: São Paulo - Escola Superior de Agricultura "Luiz de Queiroz" Publicador: São Paulo - Escola Superior de Agricultura "Luiz de Queiroz"
Tipo: Relatório
Português
Relevância na Pesquisa
38.073462%
The animal breeding values forecasting at futures times is a relevant technological innovation in the field of Animal Science, since its enables a previous indication of animals that will be either kept by the producer for breeding purposes or discarded. This study discusses an MCMC Bayesian methodology applied to panel data in a time series context. We consider Bayesian analysis of an autoregressive, AR(p), panel data model of order p, using an exact likelihood function, comparative analysis of prior distributions and predictive distributions of future observations. The methodology was tested by a simulation study using three priors: hierarchical Multivariate Normal-Inverse Gamma (model 1), independent Multivariate Student's t Inverse Gamma (model 2) and Jeffrey's (model 3). Comparisons by Pseudo-Bayes Factor favored model 2. The proposed methodology was applied to longitudinal data relative to Expected Progeny Difference (EPD) of beef cattle sires. The forecast efficiency was around 80%. Regarding the mean width of the EPD interval estimation (95%) in a future time, a great advantage was observed for the proposed Bayesian methodology over usual asymptotic frequentist method.; A previsão dos valores genéticos de animais em tempos futuros constitui importante inovação tecnológica para a área de Zootecnia...

‣ Forecasting manufacturing variation using historical process capability data : applications for random assembly, selective assembly, and serial processing

Kern, Daniel C. (Daniel Clifton), 1974-
Fonte: Massachusetts Institute of Technology Publicador: Massachusetts Institute of Technology
Tipo: Tese de Doutorado Formato: 340 p.; 18328710 bytes; 18328504 bytes; application/pdf; application/pdf
Português
Relevância na Pesquisa
17.547554%
In today's competitive marketplace, companies are under increased pressure to produce products that have a low cost and high quality. Product cost and quality are influenced by many factors. One factor that strongly influences both is manufacturing variation. Manufacturing variation is the range of values that a product's dimensions assume. Variation exists because no production process is perfect. Often times, controlling this variation is attempted during production when substantial effort and resources, e.g., time, money, and manpower, are required. The effort and resources could be reduced if the manufacturing variation could be forecast and managed during the design of the product. Traditionally, several barriers have been present that make forecasting and managing variation during the design process very challenging. The first barrier is the effort required of a design engineer to know the company's process capability, which makes it difficult to specify tolerances that can be manufactured reliably. The second barrier is the difficulty associated with understanding how a single manufacturing process or series of processes affects the variation of a product. This barrier impedes the analysis of tradeoffs among processes, the quantifying of the impact incoming stock variation has on final product variation...

‣ Bayesian analysis of autoregressive panel data model: application in genetic evaluation of beef cattle

Silva,Fabyano Fonseca e; Sáfadi,Thelma; Muniz,Joel Augusto; Rosa,Guilherme Jordão Magalhães; Aquino,Luiz Henrique de; Mourão,Gerson Barreto; Silva,Carlos Henrique Osório
Fonte: São Paulo - Escola Superior de Agricultura "Luiz de Queiroz" Publicador: São Paulo - Escola Superior de Agricultura "Luiz de Queiroz"
Tipo: Artigo de Revista Científica Formato: text/html
Publicado em 01/04/2011 Português
Relevância na Pesquisa
37.957036%
The animal breeding values forecasting at futures times is a relevant technological innovation in the field of Animal Science, since its enables a previous indication of animals that will be either kept by the producer for breeding purposes or discarded. This study discusses an MCMC Bayesian methodology applied to panel data in a time series context. We consider Bayesian analysis of an autoregressive, AR(p), panel data model of order p, using an exact likelihood function, comparative analysis of prior distributions and predictive distributions of future observations. The methodology was tested by a simulation study using three priors: hierarchical Multivariate Normal-Inverse Gamma (model 1), independent Multivariate Student's t Inverse Gamma (model 2) and Jeffrey's (model 3). Comparisons by Pseudo-Bayes Factor favored model 2. The proposed methodology was applied to longitudinal data relative to Expected Progeny Difference (EPD) of beef cattle sires. The forecast efficiency was around 80%. Regarding the mean width of the EPD interval estimation (95%) in a future time, a great advantage was observed for the proposed Bayesian methodology over usual asymptotic frequentist method.

‣ Forecasting extreme events in collective dynamics: an analytic signal approach to detecting discrete scale invariance

Viswanathan, G. M.
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Português
Relevância na Pesquisa
27.547554%
A challenging problem in physics concerns the possibility of forecasting rare but extreme phenomena such as large earthquakes, financial market crashes, and material rupture. A promising line of research involves the early detection of precursory log-periodic oscillations to help forecast extreme events in collective phenomena where discrete scale invariance plays an important role. Here I investigate two distinct approaches towards the general problem of how to detect log-periodic oscillations in arbitrary time series without prior knowledge of the location of the moveable singularity. I first show that the problem has a definite solution in Fourier space, however the technique involved requires an unrealistically large signal to noise ratio. I then show that the quadrature signal obtained via analytic continuation onto the imaginary axis, using the Hilbert transform, necessarily retains the log-periodicities found in the original signal. This finding allows the development of a new method of detecting log-periodic oscillations that relies on calculation of the instantaneous phase of the analytic signal. I illustrate the method by applying it to the well documented stock market crash of 1987. Finally, I discuss the relevance of these findings for parametric rather than nonparametric estimation of critical times.; Comment: Corrected minor mistakes...

‣ A model-free characterization of recurrences in stationary time series

Chicheportiche, Rémy; Chakraborti, Anirban
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Português
Relevância na Pesquisa
37.886245%
Study of recurrences in earthquakes, climate, financial time-series, etc. is crucial to better forecast disasters and limit their consequences. However, almost all the previous phenomenological studies involved only a long-ranged autocorrelation function, or disregarded the multi-scaling properties induced by potential higher order dependencies. Consequently, they missed the facts that non-linear dependences do impact both the statistics and dynamics of recurrence times, and that scaling arguments for the unconditional distribution may not be applicable. We argue that copulas is the correct model-free framework to study non-linear dependencies in time series and related concepts like recurrences. Fitting and/or simulating the intertemporal distribution of recurrence intervals is very much system specific, and cannot actually benefit from universal features, in contrast to the previous claims. This has important implications in epilepsy prognosis and financial risk management applications.; Comment: 4 pages, 2 figures, 2 proofs included in supplementary material

‣ Spectroscopic source redshifts and parameter constraints from weak lensing and CMB

Ishak, Mustapha; Hirata, Christopher M.
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Português
Relevância na Pesquisa
17.547554%
Weak lensing is a potentially robust and model-independent cosmological probe, but its accuracy is dependent on knowledge of the redshift distribution of the source galaxies used. The most robust way to determine the redshift distribution is via spectroscopy of a subsample of the source galaxies. We forecast constraints from combining CMB anisotropies with cosmic shear using a spectroscopically determined redshift distribution, varying the number of spectra $N_{spec}$ obtained from 64 to $\infty$. The source redshift distribution is expanded in a Fourier series, and the amplitudes of each mode are considered as parameters to be constrained via both the spectroscopic and weak lensing data. We assume independent source redshifts, and consider in what circumstances this is a good approximation (the sources are clustered and for narrow spectroscopic surveys with many objects this results in the redshifts being correlated). It is found that for the surveys considered and for a prior of 0.04 on the calibration parameters, the addition of redshift information make significant improvements on the constraints on the cosmological parameters; however, beyond $N_{spec}\sim$few$\times 10^3$ the addition of further spectra will make only a very small improvement to the cosmological parameters. We find that a better calibration makes large $N_{spec}$ more useful. Using an eigenvector analysis...

‣ Análise Bayesiana do modelo auto-regressivo para dados em painel: aplicação na avaliação genética de bovinos de corte; Bayesian analysis of autoregressive panel data model: application in genetic evaluation of beef cattle

Silva, Fabyano Fonseca e; Sáfadi, Thelma; Muniz, Joel Augusto; Rosa, Guilherme Jordão Magalhães; Aquino, Luiz Henrique de; Mourão, Gerson Barreto; Silva, Carlos Henrique Osório
Fonte: Universidade de São Paulo. Escola Superior de Agricultura Luiz de Queiroz Publicador: Universidade de São Paulo. Escola Superior de Agricultura Luiz de Queiroz
Tipo: info:eu-repo/semantics/article; info:eu-repo/semantics/publishedVersion; ; ; ; ; Formato: application/pdf
Publicado em 01/04/2011 Português
Relevância na Pesquisa
38.073462%
The animal breeding values forecasting at futures times is a relevant technological innovation in the field of Animal Science, since its enables a previous indication of animals that will be either kept by the producer for breeding purposes or discarded. This study discusses an MCMC Bayesian methodology applied to panel data in a time series context. We consider Bayesian analysis of an autoregressive, AR(p), panel data model of order p, using an exact likelihood function, comparative analysis of prior distributions and predictive distributions of future observations. The methodology was tested by a simulation study using three priors: hierarchical Multivariate Normal-Inverse Gamma (model 1), independent Multivariate Student's t Inverse Gamma (model 2) and Jeffrey's (model 3). Comparisons by Pseudo-Bayes Factor favored model 2. The proposed methodology was applied to longitudinal data relative to Expected Progeny Difference (EPD) of beef cattle sires. The forecast efficiency was around 80%. Regarding the mean width of the EPD interval estimation (95%) in a future time, a great advantage was observed for the proposed Bayesian methodology over usual asymptotic frequentist method.; A previsão dos valores genéticos de animais em tempos futuros constitui importante inovação tecnológica para a área de Zootecnia...