Made available in DSpace on 2010-04-20T20:08:30Z (GMT). No. of bitstreams: 0 Previous issue date: 1995-11-23T00:00:00Z / A metodologia para a estimação por cenários alternativos com base na interação entre modelos subjetivos causais e técnicas analíticas para o dimensionamento de mercado representa uma proposta de tratamento sistemático dos problemas de análise e previsão, utilizado nas áreas de estratégia mercadológica e planejamento empresarial. A tese trata inicialmente de dois critérios para a classificação das técnicas de previsão e geração de cenários. No primeiro deles, são considerados os atributos que usualmente orientam a seleção dos métodos de previsão mais apropriados para cada caso. No segundo, é definido um referencial com base nas regras empregadas para identificar o modo de geração dos futuros alternativos. Em ambas situações, foram consultados trabalhos que incorporaram extensivamente os conceitos de interesse desta tese, com o propósito de apoiar a sua fundamentação teórica. Uma análise adicional mais detalhada se ateve às técnicas que propiciaram a base conceitual da metodologia, a exemplo da modelagem estrutural e das probabilidades subjetivas. Com o objetivo de testar a adequação da proposta ao estudo de caso, foi escolhido o mercado atendido pela indústria de energia elétrica. Este procedimento passou inicialmente por uma investigação a nível nacional com a intenção de: (i) observar aquelas técnicas e indicadores utilizados regularmente III pelas concessionárias; (ii) constatar e comparar situações que afetam a qualidade das previsões, a exemplo do tamanho da equipe e do grau de comunicação com as instituições externas. Em seguida, foi efetuada uma aplicação prática sobre o mercado regional de eletricidade do segmento residencial, na tentativa de validar empiricamente a metodologia proposta. Como particularidades, foram introduzidas as equações simultâneas do lado analítico, e a consulta aos especialistas, os quais foram aplicados na construção do modelo exploratório do lado subjetivo. A interação entre as técnicas resultou em alguns cenários para o mercado analisado. / The methodology for forecasting by means of alternative scenarios based on the interaction between causal subjective models and analytical techniques for market measurement establishes a new framework. It provides a systematic approach to the analysis and forecasting practices in areas such as marketing strategy and business planning. We begin this thesis with a presentation of two criteria aimed at classifying the forecasting and scenario-generating techniques. As for the former, we considered some attributes which are usually applied to the selection of a suitable method under particular market conditions. Concerning the latter, we based the tipology on some rules generally employed for the specification of alternative scenarios. In both cases, to support our teorethical foundation, the research was carried out by introducing some selected techniques which strongly embody the concepts used in this thesis. Further analysis of the issue brought on some methods that defined the conceptual model, as for instance, the structural modeling and subjective probabilities. To back up the feasibility of such proposal, we chose the market supplied by the electric industry. The first step of this procedure was a national survey among electric utilities whose main purposes were: (i) point out the techniques and indicators most commonly used; (ii) consider and compare some factors that influence the quality of forecasts, such as team size and the extent of communication with external agencies. Finally, we conducted a practical application of the methodology to the regional market, as regards residential demand for electricity. As for peculiar features, we introduced in our thesis the concepts of simultaneous equations in the analytical side and the consensus analysis technique, which were applied to an exploratory model in the subjective side. The link between such techniques was used to create the scenarios for the prospective demand.

Whereas traditional data warehouse systems assume that data is complete or has been carefully preprocessed, increasingly more data is imprecise, incomplete, and inconsistent. This is especially true in the context of big data, where massive amount of data arrives continuously in real-time from vast data sources. Nevertheless, modern data analysis involves sophisticated statistical algorithm that go well beyond traditional BI and, additionally, is increasingly performed by non-expert users. Both trends require transparent data mining techniques that efficiently handle missing data and present a complete view of the database to the user. Time series forecasting estimates future, not yet available, data of a time series and represents one way of dealing with missing data. Moreover, it enables queries that retrieve a view of the database at any point in time - past, present, and future. This article presents an overview of forecasting techniques in database management systems. After discussing possible application areas for time series forecasting, we give a short mathematical background of the main forecasting concepts. We then outline various general strategies of integrating time series forecasting inside a database and discuss some individual techniques from the database community. We conclude this article by introducing a novel forecasting-enabled database management architecture that natively and transparently integrates forecast models.

The idea of time series forecasting techniques is that the past has certain information about future. So, the question of how the information is encoded in the past can be interpreted and later used to extrapolate events of future constitute the crux of time series analysis and forecasting. Several methods such as qualitative techniques (e.g., Delphi method), causal techniques (e.g., least squares regression), quantitative techniques (e.g., smoothing method, time series models) have been developed in the past in which the concept lies in establishing a model either theoretically or mathematically from past observations and estimate future from it. Of all the models, time series methods such as autoregressive moving average (ARMA) process have gained popularity because of their simplicity in implementation and accuracy in obtaining forecasts. But, these models were formulated based on certain properties that a time series is assumed to possess. Classical decomposition techniques were developed to supplement the requirements of time series models. These methods try to define a time series in terms of simple patterns called trend, cyclical and seasonal patterns along with noise. So, the idea of decomposing a time series into component patterns, later modeling each component using forecasting processes and finally combining the component forecasts to obtain actual time series predictions yielded superior performance over standard forecasting techniques. All these methods involve basic principle of moving average computation. But, the developed classical decomposition methods are disadvantageous in terms of containing fixed number of components for any time series, data independent decompositions. During moving average computation, edges of time series might not get modeled properly which affects long range forecasting. So, these issues are to be addressed by more efficient and advanced decomposition techniques such as Wavelets and Empirical Mode Decomposition (EMD). Wavelets and EMD are some of the most innovative concepts considered in time series analysis and are focused on processing nonlinear and nonstationary time series. Hence, this research has been undertaken to ascertain the predictability of nonstationary time series using wavelet and Empirical Mode Decomposition (EMD) based ARMA models. The development of wavelets has been made based on concepts of Fourier analysis and Window Fourier Transform. In accordance with this, initially, the necessity of involving the advent of wavelets has been presented. This is followed by the discussion regarding the advantages that are provided by wavelets. Primarily, the wavelets were defined in the sense of continuous time series. Later, in order to match the real world requirements, wavelets analysis has been defined in discrete scenario which is called as Discrete Wavelet Transform (DWT). The current thesis utilized DWT for performing time series decomposition. The detailed discussion regarding the theory behind time series decomposition is presented in the thesis. This is followed by description regarding mathematical viewpoint of time series decomposition using DWT, which involves decomposition algorithm. EMD also comes under same class as wavelets in the consequence of time series decomposition. EMD is developed out of the fact that most of the time series in nature contain multiple frequencies leading to existence of different scales simultaneously. This method, when compared to standard Fourier analysis and wavelet algorithms, has greater scope of adaptation in processing various nonstationary time series. The method involves decomposing any complicated time series into a very small number of finite empirical modes (IMFs-Intrinsic Mode Functions), where each mode contains information of the original time series. The algorithm of time series decomposition using EMD is presented post conceptual elucidation in the current thesis. Later, the proposed time series forecasting algorithm that couples EMD and ARMA model is presented that even considers the number of time steps ahead of which forecasting needs to be performed. In order to test the methodologies of wavelet and EMD based algorithms for prediction of time series with non stationarity, series of streamflow data from USA and rainfall data from India are used in the study. Four non-stationary streamflow sites (USGS data resources) of monthly total volumes and two non-stationary gridded rainfall sites (IMD) of monthly total rainfall are considered for the study. The predictability by the proposed algorithm is checked in two scenarios, first being six months ahead forecast and the second being twelve months ahead forecast. Normalized Root Mean Square Error (NRMSE) and Nash Sutcliffe Efficiency Index (Ef) are considered to evaluate the performance of the proposed techniques. Based on the performance measures, the results indicate that wavelet based analyses generate good variations in the case of six months ahead forecast maintaining harmony with the observed values at most of the sites. Although the methods are observed to capture the minima of the time series effectively both in the case of six and twelve months ahead predictions, better forecasts are obtained with wavelet based method over EMD based method in the case of twelve months ahead predictions. It is therefore inferred that wavelet based method has better prediction capabilities over EMD based method despite some of the limitations of time series methods and the manner in which decomposition takes place. Finally, the study concludes that the wavelet based time series algorithm could be used to model events such as droughts with reasonable accuracy. Also, some modifications that could be made in the model have been suggested which can extend the scope of applicability to other areas in the field of hydrology.