The mobile traffic in cellular networks is increasing in a steady rate as we go intoa future where we are connected to the internet practically all the time in one wayor another. To map the mobile traffic and the volume pressure on the base stationduring different time periods, it is useful to have the ability to predict the trafficvolumes within cellular networks. The data in this work consists of 4G cellular trafficdata spanning over a 7 day coherent period, collected from cells in a moderately largecity. The proposed method in this work is ARIMA modeling, in both original formand with an extension where the coefficients of the ARIMA model are re-esimated byintroducing some user characteristic variables. The re-estimated coefficients produceslightly lower forecast errors in general than a isolated ARIMA model where thevolume forecasts only depends on time. This implies that the forecasts can besomewhat improved when we allow the influence of these variables to be a part ofthe model, and not only the time series itself.

This study investigates fluctuations of construction output, and growth of the output in Kenya. Fluctuation and growth of construction activity are matters of concern in construction industries of many countries in the developing as well as in the developed world. The construction industry of Kenya is therefore an exemplifying case for this phenomenon. Construction activity in Kenya fluctuates excessively and grows very slowly. This remains a big challenge to policy makers, developers, consultants and contractors in their decision-making processes. In this study, systems thinking was applied to investigate the problem of excessive fluctuations and stunted growth of construction output in Kenya. The study developed a system dynamics model to simulate the construction output problem behaviour. The historical behaviour of the construction industry was described using construction output data of a 40-year period - from 1964 to 2003. Line graphs of the historical data exhibited profiles that helped to identify the system archetypes operating in the industry. From the profiles, it was deduced that the problem of fluctuations and slow growth of construction output in Kenya is encapsulated in two system archetypes, namely: balancing process with a delay, and limits to growth. The relationship between construction output and its determinant factors from the constru ction industry's environment was investigated using time series regression, which involved autoregressive integrated moving average (ARIMA) regression and multiple regression modelling of the output. On the basis of the historical data analysis and the system archetypes identified, a system dynamics (SD) model was developed to replicate the problem of fluctuations and slow growth in the construction output. The data used to develop the system dynamics model was annual construction output in Kenya from 1964 to 2003. The model was then used: to appraise policy changes suggested by construction industry participants in Kenya, and to project construction output in Kenya from year 2004 to year 2050, in order to establish the expected future fluctuations and growth trends of the construction output. It was observed that three fundamental changes are necessary in the system structure of the construction industry of Kenya, in order to minimize fluctuations and foster growth in construction output in the country, in the long run. The changes are: setting long-term targets of annual construction output in the industry as a whole, incorporating reserve capacity in the production process, and expanding the system st ructure to capture a larger construction market. The study recommends regulation of the response of the construction industry of Kenya to changes in construction demand in the market, and expansion of the construction industry's market into the African region and beyond.

La presente tesis doctoral se centra en el estudio biodemográfico de la mortalidad con el propósito de describir, por un lado, la evolución temporal de los patrones de mortalidad en la población de estudio y, por otra parte, analizar los factores de riesgo y los mecanismos de compensación sobre la supervivencia a diferentes edades, desde la perspectiva de la ecología humana y la biología evolutiva. Es Mercadal, en la isla de Menorca, supone un caso ideal de estudio al tratarse de una comunidad singular, formada por un grupo homogéneo de individuos, localizados en un mismo contexto social, demográfico y ecológico, que supone condiciones externas uniformes para todos los integrantes de la población. El análisis de la evolución de los patrones de mortalidad en Es Mercadal (1634-1997) pone en evidencia algunas singularidades destacables de la población relacionadas, primero, con las características ecológicas y socioeconómicas de la misma y, posteriormente, con el proceso de transición epidemiológica. A lo largo del último cuarto del siglo XIX, aparecen una serie de cambios en los patrones de mortalidad, que confirman la entrada de la población en el proceso de transición epidemiológica. La utilización de las técnicas de análisis de series temporales, los modelos ARIMA, nos han permitido analizar los patrones estacionales de las defunciones de Es Mercadal desde nuevas perspectivas. Estos modelos ponen de manifiesto la dirección de los cambios producidos en los patrones de mortalidad entre 1634 y 1997. Se aprecia una transición entre un patrón de defunciones modelable a través de un modelo ARIMA complejo, que tiene en cuenta la estacionalidad y las crisis de mortalidad recurrentes, y un modelo con un único parámetro no estacional para las series del siglo XX. Esta reducción en el número de parámetros necesarios para modelar las series temporales de defunciones coincide con el proceso de transición epidemiológica en Es Mercadal.Cuando se utiliza una aproximación biodemográfica, como es nuestro caso, los resultados nos han permitido demostrar que el análisis, en una población pretransicional, de la incidencia de la estación de nacimiento sobre la supervivencia es una herramienta efectiva para estudiar el efecto de los factores de vida temprana sobre la mortalidad a diferentes edades. Para explicar las diferencias observadas en la supervivencia a largo plazo en función de la estación de nacimiento, proponemos que el factor de vida temprana que actúa son, probablemente, las diferencias estacionales en los niveles alimenticios de la población, y, por tanto, en la nutrición de las madres. Nuestro análisis evidencia el efecto a largo plazo de los factores de vida temprana. Sin embargo, al mismo tiempo, nos hace dudar de la función adaptativa de la asociación entre las condiciones de vida durante el desarrollo y el estado de salud a edades adultas, propuesta por diferentes autores. De esta manera, los resultados no parecen apoyar la hipótesis que propone que los efectos de las condiciones de vida temprana son respuestas adaptativas del organismo. En Es Mercadal se evidencian los efectos a corto plazo de la reproducción sobre la supervivencia femenina. Este hecho puede explicarse por los costos de la reproducción sobre la salud de las mujeres, debidos a su 'fragilidad' (frailty) asociada al esfuerzo reproductor, en un ambiente donde los niveles nutricionales podían no satisfacer los requisitos energéticos necesarios. Sin embargo, no se observan evidencias de los costos de la reproducción sobre la supervivencia a edades postreproductoras. Por tanto, el mecanismo de compensación (trade-off) entre el mantenimiento y la supervivencia, se manifiesta especialmente a través de los costes a corto plazo de la reproducción. Este hecho determina la mayor mortalidad femenina a edades reproductoras durante gran parte del período en estudio. / Changes in the mortality patterns reflect both biological and sociocultural influences and are a therefore a valuable source of information on many important aspects of a population's adaptation. Death records spanning long periods of time are especially useful because of the information they can provide on the evolution of a group's ability to respond to environmental chantes.Owing to its history of relative isolation, the population of Es Mercadal In Minorca Island can be viewed as a "singular community": a small homogenous group of people living in the same social, demographic, and ecological context with more or less uniform external conditions experienced by all of its inhabitants. This homogeneity is useful from a paleodemographic standpoint because it means that variation in mortality patterns can be interpreted in terms of changes in socioeconomic and natural environmental conditions that were experienced by the entire community.In this study we have shown that ARIMA models are a useful analytical tool for exploring the complex patterns of temporal interdependence present in time series containing monthly mortality data such as those from Es Mercadal. As we have shown, this can reveal interesting long and short term trends as well as seasonal mortality patterns. Our analysis of the effects of season of birth on survival has proved to be an effective tool in determining early-life influences on mortality at different ages using a biodemographic approach in a pre-industrial setting. We explained differences in long-term survival and the probability of death at older ages according to the season of birth in terms of seasonal differences in prenatal nutritional levels. Our results provide evidence of the long-term effects of early environments, but they also raise questions about their adaptive function.We found that there were not an association between reproductive histories and long-term survival in the population studied. That fact is explained by the earlier costs of reproduction, the frailty of women associated to their reproductive effort in an environment where the nutritional levels were scarce and the infectious burden high.

Despite their success and widespread usage in industry and business, ES methods have received little attention from the statistical community. We investigate three types of statistical models that have been found to underpin ES methods. They are ARIMA models, state space models with multiple sources of error (MSOE), and state space models with a single source of error (SSOE). We establish the relationship among the three classes of models and conclude that the class of SSOE state space models is broader than the other two and provides a formal statistical foundation for ES methods. To better understand ES methods, we investigate the behaviors of ES methods for time series generated from different processes. We mainly focus on time series of ARIMA type. ES methods forecast a time series using only the series own history. To include covariates into ES methods for better forecasting a time series, we propose a new forecasting method, Exponential Smoothing with Covariates (ESCov). ESCov uses an ES method to model what left unexplained in a time series by covariates. We establish the optimality of ESCov, identify SSOE state space models underlying ESCov, and derive analytically the variances of forecasts by ESCov. Empirical studies show that ESCov outperforms ES methods and regression with ARIMA errors. We suggest a model selection procedure for choosing appropriate covariates and ES methods in practice. Computer models have been commonly used to investigate complex systems for which physical experiments are highly expensive or very time-consuming. Before using a computer model, we need to address an important question ``How well does the computer model represent the real system?" The process of addressing this question is called computer model validation that generally involves the comparison of computer outputs and physical observations. In this thesis, we propose a Bayesian approach to computer model validation. This approach integrates together computer outputs and physical observation to give a better prediction of the real system output. This prediction is then used to validate the computer model. We investigate the impacts of several factors on the performance of the proposed approach and propose a generalization to the proposed approach.

Crime is a behavior disorder that is an integrated result of social, economical and environmental factors. In the world today crime analysis is gaining significance and one of the most popular subject is crime prediction. Stakeholders of crime intend to forecast the place, time, number of crimes and crime types to get precautions. With respect to these intentions, in this thesis a spatio-temporal crime prediction model is generated by using time series forecasting with simple spatial disaggregation approach in Geographical Information Systems (GIS). The model is generated by utilizing crime data for the year 2003 in Bah&ccedil / elievler and Merkez &Ccedil / ankaya police precincts. Methodology starts with obtaining clusters with different clustering algorithms. Then clustering methods are compared in terms of land-use and representation to select the most appropriate clustering algorithms. Later crime data is divided into daily apoch, to observe spatio-temporal distribution of crime. In order to predict crime in time dimension a time series model (ARIMA) is fitted for each week day, Then the forecasted crime occurrences in time are disagregated according to spatial crime cluster patterns. Hence the model proposed in this thesis can give crime prediction in both space and time to help police departments in tactical and planning operations.

Katholische Universiẗat, Diss.--Eichstätt-Ingolstadt, 2006.

Foreign exchange (FX) rate forecasting has been a challenging area of study in the past. Various linear and nonlinear methods have been used to forecast FX rates. As the currency data are nonlinear and highly correlated, forecasting through nonlinear dynamical systems is becoming more relevant. The nearest neighbor (NN) algorithm is one of the most commonly used nonlinear pattern recognition and forecasting methods that outperforms the available linear forecasting methods for the high frequency foreign exchange data. The basic idea behind the NN is to capture the local behavior of the data by selecting the instances having similar dynamic behavior. The most relevant k number of histories to the present dynamical structure are the only past values used to predict the future. Due to this reason, NN algorithm is also known as the k-nearest neighbor algorithm (k-NN). Here k represents the number of chosen neighbors. In the k-nearest neighbor forecasting procedure, similar instances are captured through a distance function. Since the forecasts completely depend on the chosen nearest neighbors, the distance plays a key role in the k-NN algorithm. By choosing an appropriate distance, we can improve the performance of the algorithm significantly. The most commonly used distance for k-NN forecasting in the past was the Euclidean distance. Due to possible correlation among vectors at different time frames, distances based on deterministic vectors, such as Euclidean, are not very appropriate when applying for foreign exchange data. Since Mahalanobis distance captures the correlations, we suggest using this distance in the selection of neighbors. In the present study, we used five different foreign currencies, which are among the most traded currencies, to compare the performances of the k-NN algorithm with traditional Euclidean and Absolute distances to performances with the proposed Mahalanobis distance. The performances were compared in two ways: (i) forecast accuracy and (ii) transforming their forecasts in to a more effective technical trading rule. The results were obtained with real FX trading data, and the results showed that the method introduced in this work outperforms the other popular methods. Furthermore, we conducted a thorough investigation of optimal parameter choice with different distance measures. We adopted the concept of distance based weighting to the NN and compared the performances with traditional unweighted NN algorithm based forecasting. Time series forecasting methods, such as Auto regressive integrated moving average process (ARIMA), are widely used in many ares of time series as a forecasting technique. We compared the performances of proposed Mahalanobis distance based k-NN forecasting procedure with the traditional general ARIM- based forecasting algorithm. In this case the forecasts were also transformed into a technical trading strategy to create buy and sell signals. The two methods were evaluated for their forecasting accuracy and trading performances. Multi-step ahead forecasting is an important aspect of time series forecasting. Even though many researchers claim that the k-Nearest Neighbor forecasting procedure outperforms the linear forecasting methods for financial time series data, and the available work in the literature supports this claim with one step ahead forecasting. One of our goals in this work was to improve FX trading with multi-step ahead forecasting. A popular multi-step ahead forecasting strategy was adopted in our work to obtain more than one day ahead forecasts. We performed a comparative study on the performance of single step ahead trading strategy and multi-step ahead trading strategy by using five foreign currency data with Mahalanobis distance based k-nearest neighbor algorithm.

A capacidade de prever resultados futuros, ao se analisar uma série de dados, é uma importante ferramenta para o planejamento de qualquer empresa ou indústria. Porém, a literatura oferece muitas opções de ferramentas e modelos estatísticos que permitem obter estas previsões. Cada qual com suas características e recomendações. Dentre estes modelos, destacam-se os modelos de Box e Jenkins, e os modelos de Redes Neurais Artificiais (RNA) - com destaque aos modelos de perceptron de múltiplas camadas (MLP). Estas duas diferentes abordagens são comparadas nesta dissertação com relação a sua capacidade de obter previsões acuradas em séries de dados não lineares quanto a sua média. As abordagens foram comparadas utilizando-se a série mensal do índice de produção física industrial do Estado do Rio Grande do Sul. Bem como a série anual de manchas solares, sendo a segunda utilizada como caso-controle para as comparações, devido ao fato de que as suas propriedades já foram amplamente estudadas. No estudo da série do índice de produção física mensal, os modelos de Box e Jenkins obtiveram melhor rendimento. Na série das manchas solares foram os modelos MLP que se destacaram. Desta forma, não é possível afirmar se alguma das abordagens é superior - tratando-se de séries de dados não lineares quanto a sua média. / The capacity to preview future outcomes on the time series analysis is an important tool for any business and industry planning. However, the literature offers many options on statistical tools and models which allow to obtain these forecasts. Each one with their features and recommendations. 1n these models, the Box and Jenkins and Artificial Neural Networks (ANN) models, with the multilayer perceptron (MLP) highlighted, stand out. These two different approaches are compared in this thesis related to the capacity to obtain accurate forecasts in mean related non-linear time series analysis. These approaches were compared using the monthly physical production index of Rio Grande do Sul time series and the sunspot series, being the second one used as a case-control to the comparisons, due the fact of its properties are already widely studied. 1n the monthly physical production index series study, t,he Box and Jenkins models obtained better efficiency. 1n the sunspot series, the MLP models were highlighted. So, it isn't possible to affirm if any of the approaches is superior, in the case of mean related non-linear time series.

A capacidade de prever resultados futuros, ao se analisar uma série de dados, é uma importante ferramenta para o planejamento de qualquer empresa ou indústria. Porém, a literatura oferece muitas opções de ferramentas e modelos estatísticos que permitem obter estas previsões. Cada qual com suas características e recomendações. Dentre estes modelos, destacam-se os modelos de Box e Jenkins, e os modelos de Redes Neurais Artificiais (RNA) - com destaque aos modelos de perceptron de múltiplas camadas (MLP). Estas duas diferentes abordagens são comparadas nesta dissertação com relação a sua capacidade de obter previsões acuradas em séries de dados não lineares quanto a sua média. As abordagens foram comparadas utilizando-se a série mensal do índice de produção física industrial do Estado do Rio Grande do Sul. Bem como a série anual de manchas solares, sendo a segunda utilizada como caso-controle para as comparações, devido ao fato de que as suas propriedades já foram amplamente estudadas. No estudo da série do índice de produção física mensal, os modelos de Box e Jenkins obtiveram melhor rendimento. Na série das manchas solares foram os modelos MLP que se destacaram. Desta forma, não é possível afirmar se alguma das abordagens é superior - tratando-se de séries de dados não lineares quanto a sua média. / The capacity to preview future outcomes on the time series analysis is an important tool for any business and industry planning. However, the literature offers many options on statistical tools and models which allow to obtain these forecasts. Each one with their features and recommendations. 1n these models, the Box and Jenkins and Artificial Neural Networks (ANN) models, with the multilayer perceptron (MLP) highlighted, stand out. These two different approaches are compared in this thesis related to the capacity to obtain accurate forecasts in mean related non-linear time series analysis. These approaches were compared using the monthly physical production index of Rio Grande do Sul time series and the sunspot series, being the second one used as a case-control to the comparisons, due the fact of its properties are already widely studied. 1n the monthly physical production index series study, t,he Box and Jenkins models obtained better efficiency. 1n the sunspot series, the MLP models were highlighted. So, it isn't possible to affirm if any of the approaches is superior, in the case of mean related non-linear time series.

Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-07-16T19:54:29Z No. of bitstreams: 1 ARELI MESQUITA DA SILVA - DISSERTAÇÃO PPGMAT 2007..pdf: 701919 bytes, checksum: 78ea7b65513f1fe6d83acdb4f3030b43 (MD5) / Made available in DSpace on 2018-07-16T19:54:29Z (GMT). No. of bitstreams: 1 ARELI MESQUITA DA SILVA - DISSERTAÇÃO PPGMAT 2007..pdf: 701919 bytes, checksum: 78ea7b65513f1fe6d83acdb4f3030b43 (MD5) Previous issue date: 2007-07 / Séries temporais envolvendo dados angulares aparecem nas mais diversas áreas do conhecimento. Por exemplo, na perfuração de um poço petrolífero direcional, o deslocamento da broca de perfuração, ao longo da trajetória do poço, pode ser considerado uma realização de uma série temporal de dados angulares. Um dos interesses, neste contexto, consiste em realizar previsões de posicionamentos futuros da broca de perfuração, as quais darão mais apoio ao engenheiro de petróleo na tomada de decisão de quando e como interferir na trajetória de um poço, de modo que este siga o curso planejado. Neste trabalho, estudamos algumas classes de modelos que podem ser utilizados para a modelagem desse tipo de série. / Time series involving angular data appear in many diverse areas of scientific knowledge. For example, in the drilling of a directional oil well, the displacement of the drill, along the path of the well, can be considered as an angular data time series. One of the objectives, in this context, consists in carrying out forecasts of the future positions of the drill, which will give more support to the petroleum engineer in the decision-making of when and how interfere in the path of a well, so that this follows the planned course. In this work, we study some classes of models that can be utilized for the modeling of that kind of series.