Papers by Rosangela Ballini
Forecasting Of streamflows by monthly averages using neural fuzzy networks
This paper presents a neural fuzzy network model for seasonal streamflow forecasting. The model i... more This paper presents a neural fuzzy network model for seasonal streamflow forecasting. The model is based on a constructive learning method where neurons groups compete when the network receives a new input, so that it learns the fuzzy rules and membership functions essential for modelling a fuzzy system. The model was applied to the problem of seasonal streamflow forecasting using a database of average monthly inflows of three Brazilian hydroelectric plants located at different river basins. The performance of the model developed was compared with conventional approaches used to forecast streamflows. The results show that the neural fuzzy network model provides a better one-step-ahead streamflow forecasting, with forecasting errors significantly lower than the other approaches.Este trabalho apresenta um modelo de rede neural nebulosa para previsão de vazões sazonais. O modelo é baseado em um método de aprendizado construtivo onde grupos de neurônios competem quando a rede recebe uma...
REGEPE - Revista de Empreendedorismo e Gestão de Pequenas Empresas, 2021
Objective: We aim to answer four questions. First, with the increasing number of publications, is... more Objective: We aim to answer four questions. First, with the increasing number of publications, is there a concentration in specific subjects, or on the contrary, a dispersion, amplifying the span of themes related to entrepreneurship? Second, is there a hierarchy of subjects, in the sense that some of them constitute the “core” of entrepreneurship? Third, are they connected with other established research areas? Finally, it is possible to identify papers that are influential, acting as hubs in the cluster’s formation? Method: We developed an original version of the computational procedure proposed by Shibata et al (2008), which allows us to understand the diversity of the different sub-areas of the topic investigated, reducing the need for specialist supervision. Originality / Relevance: We developed and applied a method to capture the formation and evolution of research areas in entrepreneurship literature, via direct citation networks, allowing us to understand the iteration betwe...
Knowledge-Based Systems, 2018
This paper suggests a fuzzy inference system (iFIS) modeling approach for interval-valued time se... more This paper suggests a fuzzy inference system (iFIS) modeling approach for interval-valued time series forecasting. Interval-valued data arise quite naturally in many situations in which such data represent uncertainty/variability or when comprehensive ways to summarize large data sets are required. The method comprises a fuzzy rule-based framework with affine consequents which provides a (non)linear framework that processes interval-valued symbolic data. The iFIS antecedents identification uses a fuzzy c-means clustering algorithm for interval-valued data with adaptive distances, whereas parameters of the linear consequents are estimated with a center-range methodology to fit a linear regression model to symbolic interval data. iFIS forecasting power, measured by accuracy metrics and statistical tests, was evaluated through Monte Carlo experiments using both synthetic interval-valued time series with linear and chaotic dynamics, and real financial interval-valued time series. The results indicate a superior performance of iFIS compared to traditional alternative single-valued and interval-valued forecasting models by reducing 19% on average the predicting errors, indicating that the suggested approach can be considered as a promising tool for interval time series forecasting.
Anais do 10. Congresso Brasileiro de Inteligência Computacional, 2016
Resumo-Este artigo apresenta os resultados de prospecção de um e doze passos à frente da demanda ... more Resumo-Este artigo apresenta os resultados de prospecção de um e doze passos à frente da demanda mensal de energia elétrica de uma concessionária de energia pertencente a região sudeste do Brasil. Neste trabalho a demanda de energia elétrica total é subdividida em três grupos de consumo: residencial, industrial e comercial. O modelo de previsão adotado é baseado em regras nebulosas do tipo Takagi-Sugeno (TS), sendo o número de regras obtido via algoritmo de agrupamento não supervisionado Subtractive Clustering. Uma base de regras nebulosa é determinada para cada classe de consumo e os parâmetros do sistema de inferência são ajustados usando o algoritmo de otimização de maximização da verossimilhança. Como variáveis de entrada são consideradas as observações de demanda em instantes anteriores além de variáveis explicativas de natureza macroeconômica. O desempenho do modelo é verificado por meio de medidas de erros calculadas dentro e fora da amostra e os resultados indicam que o sistema de inferência nebuloso atinge índices de desempenho na ordem anual de 3% para as classes de consumo. Palavras-chave-Demanda de energia elétrica, sistema de inferência nebuloso, previsão, séries temporais.
Previs�o de vaz�es m�dias mensais usando redes neurais nebulosas
Anais do 11. Congresso Brasileiro de Inteligência Computacional, 2016
Evolving Systems, 2014
In this paper we propose a novel approach for time series forecasting based on ordered weighted a... more In this paper we propose a novel approach for time series forecasting based on ordered weighted averaging operators (OWA) as linear filter and forecasting models. The OWA operators describe a family of averaging operators parameterized by the choice the weights or filter coefficients. Starting with a nonstationary time series of a given phenomenon, we evaluate the use of linear decaying and constant weights as filtering processes of time series. Moreover, we investigate the effectiveness of using exponential weighted moving average process as a filter linear. After the application of the linear operators, we formulate the best possible forecasting models, ARIMA and neural network models, for short-term forecasting for each of the new structured time series using the usual optimal procedures for a real load data from the Southest Brazilian Company. A residual analysis of these forecasting models is given. In addition, the classical ARIMA and neural network models are also developed for the subject data, and the results are compared with the proposed models that we have introduced. In all cases the new models give better short-term forecasting results than the classical models.
Revista DAE, 2009
DAE maio/09 RESUMO O objetivo do artigo foi avaliar os impactos econômicos diretos e indiretos so... more DAE maio/09 RESUMO O objetivo do artigo foi avaliar os impactos econômicos diretos e indiretos sobre a economia brasileira dos investimentos no setor de saneamento básico, a partir da matriz de insumo-produto. Embora o SSB represente apenas 0.59% do PIB brasileiro, sua capacidade de encadeamento produtivo e de geração de renda e emprego dentro e fora do setor é bastante elevada. Para cada R$ 1 bilhão de investimento no setor teríamos: a) um aumento de R$ 1,7 bilhão no valor da produção da economia; b) uma expansão de R$ 245 milhões da massa salarial, de R$ 355 milhões do excedente operacional bruto e de R$ 139 milhões em impostos diretos e indiretos; e c) a geração de 42 mil novos empregos diretos e indiretos em toda cadeia produtiva.
Linear Decaying Weights for Time Series Smoothing: An Analysis
International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems, 2014
In this paper, we investigate the use of weighted averaging aggregation operators as techniques f... more In this paper, we investigate the use of weighted averaging aggregation operators as techniques for time series smoothing. We analyze the moving average, exponential smoothing methods, and a new class of smoothing operators based on linearly decaying weights from the perspective of ordered weights averaging to estimate a constant model. We examine two important features associated with the smoothing processes: the average age of the data and the expected variance, both defined in terms of the associated weights. We show that there exists a fundamental conflict between keeping the variance small while using the freshest data. We illustrate the flexibility of the smoothing methods with real datasets; that is, we evaluate the aggregation operators with respect to their minimal attainable variance versus average age. We also examine the efficiency of the smoothed models in time series smoothing, considering real datasets. Good smoothing generally depends upon the underlying method's...
Anais do 7. Congresso Brasileiro de Redes Neurais, 2016
Resumo-O presente trabalho aplica um modelo baseado em agrupamento nebuloso no problema de previs... more Resumo-O presente trabalho aplica um modelo baseado em agrupamento nebuloso no problema de previsão de carga de curto prazo para as 24 horas do próximo dia. Neste modelo,é utilizado o algoritmo de agrupamento fuzzy c-means para explorar a estrutura dos dados históricos, e reconhecimento de padrões para capturar similaridades na tendência das séries de consumo de carga discretizadas em base horária. A metodologiá e testada em um ponto de medição tipicamente residencial localizado na região nordeste do Brasil. Os resultados mostram erros percentuais absolutos médios na ordem de 1,99%.
Innovations and Solutions
The main objective of this chapter is to present a hybrid model for bus load forecasting. This ap... more The main objective of this chapter is to present a hybrid model for bus load forecasting. This approach represents an essential tool for the operation of the electrical power system and the hybrid model combines a bus clustering process and a load forecasting model. As a case study, the model was applied to the real Brazilian electrical system, and the results revealed a performance similar to that of conventional models for bus load forecasting, but about 14 times faster. The results are compatible with the safe operating load levels for the Brazilian electrical power system and have proved to be adequate for use in real operation tasks.
An hybrid aggregate model applied to the short-term bus load forecasting problem
2009 IEEE Bucharest PowerTech, 2009
AbstractIn this paper we present a hybrid methodology built on a combination of clustering and f... more AbstractIn this paper we present a hybrid methodology built on a combination of clustering and forecasting techniques used to solve the short-term bus load forecasting problem. The proposed method was made in two phases: In the first phase a clustering algorithm is used to ...
The 2006 IEEE International Joint Conference on Neural Network Proceedings, 2006
In this paper we present a methodology based on a combination of many distinct predictors in an e... more In this paper we present a methodology based on a combination of many distinct predictors in an ensemble, named hybrid ensemble model, to obtain a more accurate output using the results of single predictors. As basic components, we have used Artificial Neural Networks and Support Vector Machines models. In order to evaluate the performance, the hybrid model was required to predict a 24h daily series energy consumption of a Brazilian electrical operation unit located in the northeast of Brazil. The proposed ensemble model has reached an error 25% smaller than that achieved by the best single predictor. The model was initialized several times to confirm that ensembles of predictors also tend to produce low variance profiles.
2010 10th International Conference on Hybrid Intelligent Systems, 2010
This paper proposes a methodology for short-term bus load forecasting. This approach calculates t... more This paper proposes a methodology for short-term bus load forecasting. This approach calculates the short-term bus load forecast using few aggregate models. The idea is to cluster the buses in groups with similar daily load profiles and for each cluster an aggregate forecasting model is built based on the analysis of individual bus load data. The solution obtained through aggregate approach is similar to the solution obtained by individual bus load forecasting model, but with lower computational effort. This proposed methodology was implemented in a user friendly computational forecasting support system. The use of the computational prediction was essential to provide ease and speed in finding the solutions.
An aggregate model applied to the short-term bus load forecasting problem
2009 IEEE/PES Power Systems Conference and Exposition, 2009
AbstractIn this paper we present a methodology based on a combination of clustering and forecast... more AbstractIn this paper we present a methodology based on a combination of clustering and forecasting techniques. The proposed method is built in two phases: In the first phase, a clustering algorithm is used to identify buses clusters with similar daily load profile. In the second ...
Ensembles of Selected and Evolved Predictors using Genetic Algorithms for Time Series Prediction
2006 IEEE International Conference on Evolutionary Computation
This work proposes the use of Neural Networks Ensembles to predict future values of an electrical... more This work proposes the use of Neural Networks Ensembles to predict future values of an electrical load time series. At first, to generate these ensembles it is necessary to make several predictions of the same time series using various different networks in which every single one alone is sufficiently competent to predict the above mentioned time series. Therefore, we applied
Clustering Bus Load Curves
IEEE PES Power Systems Conference and Exposition, 2004.
This paper proposes the clustering of a set of busses through a fuzzy c-means clustering approach... more This paper proposes the clustering of a set of busses through a fuzzy c-means clustering approach. The utilization of fuzzy techniques in a clustering problem aims the attainment of a partition fuzzy in the data set, allowing degrees of relationship between different elements of the set, this way, an element can belong to more than one group with different membership
Agrupamento de curvas de demanda
Proceedings of CBA-Congresso …, 2004
Abstract This paper proposes the clustering of a set of bars through a fuzzy c-means clustering ... more Abstract This paper proposes the clustering of a set of bars through a fuzzy c-means clustering approach. The clustering algorithm aims at exploring similar data features and determining which bars of the system that presents a consumption profile analogous to a specific weekday. ...
Market risk exposure plays a key role for financial institutions risk management. A possible meas... more Market risk exposure plays a key role for financial institutions risk management. A possible measure for this exposure is to evaluate losses likely to incur when the price of the portfolio’s assets declines using Value-at-Risk (VaR) estimates, one of the most prominent measure of financial downside market risk. This paper suggests an evolving possibilistic fuzzy modeling approach for VaR estimation. The approach is based on an extension of the possibilistic fuzzy c-means clustering and functional fuzzy rule-based modeling, which employs memberships and typicalities to update clusters and creates new clusters based on a statistical control distance-based criteria. ePFM also uses an utility measure to evaluate the quality of the current cluster structure. Computational experiments consider data of the main equity market index of Brazil, Ibovespa, from January 2000 to December 2012 for VaR estimation using ePFM, traditional econometric benchmarks such as GARCH and EWMA, and state of th...
European Society for Fuzzy Logic and Technology, 2007
This paper introduces a new approach to ad- just a class of neurofuzzy networks based on the idea... more This paper introduces a new approach to ad- just a class of neurofuzzy networks based on the idea of participatory learning. Partici- patory learning is a mean to learn and revise beliefs based on what is already known or be- lieved. The performance of the approach is veried with the Box and Jenkins gas fur- nace modeling problem, and with
Uploads
Papers by Rosangela Ballini