Previsão do índice bovespa por meio de redes neurais artificiais: uma análise comparada aos métodos tradicionais de séries de tempo

Forecast is the basis for making strategic, tactical and operational business decisions. In financial economics, several techniques have been used to predict the behavior of assets over the past decades.Thus, there are several methods to assist in the task of time series forecasting, however, conventional modeling techniques such as statistical models and those based on theoretical mathematical models have produced unsatisfactory predictions, increasing the number of studies in more advanced methods of prediction. Among these, the Artificial Neural Networks (ANN) are a relatively new and promising method for predicting business that shows a technique that has caused much interest in the financial environment and has been used successfully in a wide variety of financial modeling systems applications, in many cases proving its superiority over the statistical models ARIMA-GARCH. In this context, this study aimed to examine whether the ANNs are a more appropriate method for predicting the behavior of Indices in Capital Markets than the traditional methods of time series analysis. For this purpose we developed an quantitative study, from financial economic indices, and developed two models of RNA-type feedfoward supervised learning, whose structures consisted of 20 data in the input layer, 90 neurons in one hidden layer and one given as the output layer (Ibovespa). These models used backpropagation, an input activation function based on the tangent sigmoid and a linear output function. Since the aim of analyzing the adherence of the Method of Artificial Neural Networks to carry out predictions of the Ibovespa, we chose to perform this analysis by comparing results between this and Time Series Predictive Model GARCH, developing a GARCH model (1.1).Once applied both methods (ANN and GARCH) we conducted the results' analysis by comparing the results of the forecast with the historical data and by studying the forecast errors by the MSE, RMSE, MAE, Standard Deviation, the Theil's U and forecasting encompassing tests. It was found that the models developed by means of ANNs had lower MSE, RMSE and MAE than the GARCH (1,1) model and Theil U test indicated that the three models have smaller errors than those of a naïve forecast. Although the ANN based on returns have lower precision indicator values than those of ANN based on prices, the forecast encompassing test rejected the hypothesis that this model is better than that, indicating that the ANN models have a similar level of accuracy . It was concluded that for the data series studied the ANN models show a more appropriate Ibovespa forecasting than the traditional models of time series, represented by the GARCH model

Pedagogia Freinet: a construção de uma práxis em turmas de 5ª a 8ª séries

Ce travail que nous venons de présenter apport une systématisation sur la viabilité de la praxis de l´éducateur français Célestin Freinet (1896-1996). Il s´agit d´une étude qualificative, recherche-action, développée dans une école coopérative l´ École Freinet à Natal/RN Brésil, auprès des élèves du 5ème et 8ème séries de l´Enseignement Fondamental. Le problème central que nous a amené à cette recherche, a été le fait de vérifier les progressions dans la production scientifique et dans les pratiques des enseignants qui ont à la charge l´Éducation Infantine et les premières séries de l´Enseignement Fondamental. Tout cela en tenant compte en concevoir l´élève en tant que sujet actif de son apprentissage et coresponsable à l´organisation du travail scolaire. Toutefois, qaund-il s´agit des quatre dernières séries de l´Enseignement Fondamental (5ème au 8ème séries), les propositions progressistes sont prises comme quelque chose pratiquement impossible d´en mettre en marche. Pour attenuer la lacune théorique sur le sujet, nous avons offert un texte significatif pour les enseignants d´une façon générale, les chercheurs de l´académie et pour l´écolechamp de la recherché. Tout cela par le fait de ne pas donner de priorité à une technique ou à un principe, en particulier. Cette pédagogie est discutée autant que possible dans son contexte global et aussi sur le comment s´est manifestée dans sa práxis . Nous avons mis en relief les aspects importants pour le fonctionnement de la salle de classe, aussi que les estratégies utilisées pour la construction, communication et documentation des connaissances L´intervention nous a permis de construire l´objet par la réflexion en action, à partir des données obtenues pendant les observations, les entretiens, les enterviews, les moments d´études et de la quête des textes écrits. L´analyse a été faite à la lumière de la methodologie comprehensive (Kaufmann, 1996), entrelacée au materiau empirique et à la théorie. Cette forme d´analyse nous a permis d´obtenir la logique dans l´ensemble comme nous crayons que doit caracteriser la recherche aux Sciences Humaines et Sociales. Les considérations finales mettent l´accent d´un côté sur l´efficacité de la pédagogie Freinet, pour orienter le processus de l´enseignement apprentissage, en tous les dégres de l´enseignement et de l´autre côté nous montre la nécessité urgente de l´équipe de l´école champ de la recherche, de qualifier sa pratique à partir des références théoriques-pratiques de la pédagogie Freinet

Reflexões sobre a construção e evolução de conceitos geométricos nas séries intermediárias do ensino fundamental

This thesis represents a didactic research linked to the Post-graduation Programme in Education of the Universidade Federal do Rio Grande do Norte which aimed to approach the construction of the geometrical concepts of Volume of the Rectangular Parallelepiped, Area and Perimeter of the Rectangle adding a study of the Area of the Circle. The research was developed along with students from the 6th level of the Elementary School, in a public school in Natal/RN. The pedagogical intervention was made up of three moments: application of a diagnostic evaluation, instrument that enabled the creation of the teaching module by showing the level of the geometry knowledge of the students; introduction of a Teaching Module by Activities aiming to propose a reflexive didactic routing directed to the conceptual construction because we believed that such an approach would favor the consolidation of the learning process by becoming significant to the apprentice, and the accomplishment of a Final Evaluation through which we established a comparison of the results obtained before and after the teaching intervention. The data gathered were analyzed qualitatively by means of a study of understanding categories of mathematical concepts, in addition to using descriptive statistics under the quantitative aspect. Based on the theory of Richard Skemp, about categorization of mathematical knowledge, in the levels of Relational and Instrumental Understanding were achieved in contextual situations and varied proportions, thus enabling a contribution in the learning of the geometrical concepts studied along with the students who took part in the research. We believe that this work may contribute with reflections about the learning processes, a concern which remained during all the stages of the research, and also that the technical competence along with the knowledge about the constructivist theory will condition the implementation of a new dynamics to the teaching and learning processes. We hope that the present research work may add some contribution to the teaching practice in the context of the teaching of Mathematics for the intermediate levels of the Elementary School

Extração e Representação de Conhecimento de Séries Temporais de Demanda de Energia Elétrica Usando TSKR

The opening of the Brazilian market of electricity and competitiveness between companies in the energy sector make the search for useful information and tools that will assist in decision making activities, increase by the concessionaires. An important source of knowledge for these utilities is the time series of energy demand. The identification of behavior patterns and description of events become important for the planning execution, seeking improvements in service quality and financial benefits. This dissertation presents a methodology based on mining and representation tools of time series, in order to extract knowledge that relate series of electricity demand in various substations connected of a electric utility. The method exploits the relationship of duration, coincidence and partial order of events in multi-dimensionals time series. To represent the knowledge is used the language proposed by Mörchen (2005) called Time Series Knowledge Representation (TSKR). We conducted a case study using time series of energy demand of 8 substations interconnected by a ring system, which feeds the metropolitan area of Goiânia-GO, provided by CELG (Companhia Energética de Goiás), responsible for the service of power distribution in the state of Goiás (Brazil). Using the proposed methodology were extracted three levels of knowledge that describe the behavior of the system studied, representing clearly the system dynamics, becoming a tool to assist planning activities

Estudos de modelos Estocásticos: processo de contato, séries aleatórias e processos difusivos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Predição de séries temporais de parâmetros de redes HSPA - WCDMA

With the growing demand of data traffic in the networks of third generation (3G), the mobile operators have attempted to focus resources on infrastructure in places where it identifies a greater need. The channeling investments aim to maintain the quality of service especially in dense urban areas. WCDMA - HSPA parameters Rx Power, RSCP (Received Signal Code Power), Ec/Io (Energy per chip/Interference) and transmission rate (throughput) at the physical layer are analyzed. In this work the prediction of time series on HSPA network is performed. The collection of values of the parameters was performed on a fully operational network through a drive test in Natal - RN, a capital city of Brazil northeastern. The models used for prediction of time series were the Simple Exponential Smoothing, Holt, Holt Winters Additive and Holt Winters Multiplicative. The objective of the predictions of the series is to check which model will generate the best predictions of network parameters WCDMA - HSPA.

Análise de métodos para detecção de mudanças estruturais em séries temporais climáticas

The time series analysis has played an increasingly important role in weather and climate studies. The success of these studies depends crucially on the knowledge of the quality of climate data such as, for instance, air temperature and rainfall data. For this reason, one of the main challenges for the researchers in this field is to obtain homogeneous series. A time series of climate data is considered homogeneous when the values of the observed data can change only due to climatic factors, i.e., without any interference from external non-climatic factors. Such non-climatic factors may produce undesirable effects in the time series, as unrealistic homogeneity breaks, trends and jumps. In the present work it was investigated climatic time series for the city of Natal, RN, namely air temperature and rainfall time series, for the period spanning from 1961 to 2012. The main purpose was to carry out an analysis in order to check the occurrence of homogeneity breaks or trends in the series under investigation. To this purpose, it was applied some basic statistical procedures, such as normality and independence tests. The occurrence of trends was investigated by linear regression analysis, as well as by the Spearman and Mann-Kendall tests. The homogeneity was investigated by the SNHT, as well as by the Easterling-Peterson and Mann-Whitney-Pettit tests. Analyzes with respect to normality showed divergence in their results. The von Neumann ratio test showed that in the case of the air temperature series the data are not independent and identically distributed (iid), whereas for the rainfall series the data are iid. According to the applied testings, both series display trends. The mean air temperature series displays an increasing trend, whereas the rainfall series shows an decreasing trend. Finally, the homogeneity tests revealed that all series under investigations present inhomogeneities, although they breaks depend on the applied test. In summary, the results showed that the chosen techniques may be applied in order to verify how well the studied time series are characterized. Therefore, these results should be used as a guide for further investigations about the statistical climatology of Natal or even of any other place.

Previsão do índice bovespa por meio de redes neurais artificiais: uma análise comparada aos métodos tradicionais de séries de tempo

Forecast is the basis for making strategic, tactical and operational business decisions. In financial economics, several techniques have been used to predict the behavior of assets over the past decades.Thus, there are several methods to assist in the task of time series forecasting, however, conventional modeling techniques such as statistical models and those based on theoretical mathematical models have produced unsatisfactory predictions, increasing the number of studies in more advanced methods of prediction. Among these, the Artificial Neural Networks (ANN) are a relatively new and promising method for predicting business that shows a technique that has caused much interest in the financial environment and has been used successfully in a wide variety of financial modeling systems applications, in many cases proving its superiority over the statistical models ARIMA-GARCH. In this context, this study aimed to examine whether the ANNs are a more appropriate method for predicting the behavior of Indices in Capital Markets than the traditional methods of time series analysis. For this purpose we developed an quantitative study, from financial economic indices, and developed two models of RNA-type feedfoward supervised learning, whose structures consisted of 20 data in the input layer, 90 neurons in one hidden layer and one given as the output layer (Ibovespa). These models used backpropagation, an input activation function based on the tangent sigmoid and a linear output function. Since the aim of analyzing the adherence of the Method of Artificial Neural Networks to carry out predictions of the Ibovespa, we chose to perform this analysis by comparing results between this and Time Series Predictive Model GARCH, developing a GARCH model (1.1).Once applied both methods (ANN and GARCH) we conducted the results' analysis by comparing the results of the forecast with the historical data and by studying the forecast errors by the MSE, RMSE, MAE, Standard Deviation, the Theil's U and forecasting encompassing tests. It was found that the models developed by means of ANNs had lower MSE, RMSE and MAE than the GARCH (1,1) model and Theil U test indicated that the three models have smaller errors than those of a naïve forecast. Although the ANN based on returns have lower precision indicator values than those of ANN based on prices, the forecast encompassing test rejected the hypothesis that this model is better than that, indicating that the ANN models have a similar level of accuracy . It was concluded that for the data series studied the ANN models show a more appropriate Ibovespa forecasting than the traditional models of time series, represented by the GARCH model