A previsão de demanda é uma das principais ferramentas para a eficiência do gerenciamento das organizações, afetando diretamente a lucratividade do negócio. O atual nível competitivo das empresas requer previsões cada vez mais acuradas, sedo estas um diferencial para o sucesso empresarial. Neste contexto, a combinação de previsões se tornou um dos principais métodos empregados no intuito de melhorar a precisão das previsões. Através de uma revisão da literatura sobre as abordagens da combinação de previsões, identificou-se uma carência de estudos comparativos que incorporem modelos de regressão para a combinação de previsões. Assim, o objetivo principal desta dissertação é combinar três previsões individuais (redes neurais, modelos ARIMA e modelos de alisamento exponencial) via média simples, variância mínima e modelos de regressão, comparando as três previsões combinadas com suas previsões individuais. Estas comparações serão avaliadas em duas situações: em séries simuladas (estacionárias) e em uma série de dados reais (não estacionária) de uma empresa que realiza auditorias médicas. As medidas empregadas para a escolha do método mais preciso são MAE, MAPE, RMSE e o coeficiente U de Theil. Os resultados obtidos enfatizam a melhoria das previsões quando estas são combinadas por regressão, tanto para séries convergentes quanto para a série divergente. / Forecasting is a key tool for ensuring the efficiency of management in organizations, directly affecting business profitability. The current competitive corporative level requires increasingly accurate predictions. In this context, the combination of forecasts has improved forecast accuracy. Through a literature review on the approaches of combining forecasts, we identified a lack of comparative studies that incorporate regression models for combining forecasts. Thus, the main objective of this dissertation is to combine three individual forecasts (neural networks, ARIMA models and exponential smoothing models) via simple average, minimum variance and regression models, comparing the three combined forecasts with their individual forecasts. These comparisons are evaluated in two situations: in simulated series (converging) and in series of real data (divergent) from a company that performs medical audits. The measures used to identify the best method are MAE, MAPE, RMSE and Theil’s U coefficient. Results from combined methods improved the predictions in both convergent and divergent series.

A previsão de demanda é uma das principais ferramentas para a eficiência do gerenciamento das organizações, afetando diretamente a lucratividade do negócio. O atual nível competitivo das empresas requer previsões cada vez mais acuradas, sedo estas um diferencial para o sucesso empresarial. Neste contexto, a combinação de previsões se tornou um dos principais métodos empregados no intuito de melhorar a precisão das previsões. Através de uma revisão da literatura sobre as abordagens da combinação de previsões, identificou-se uma carência de estudos comparativos que incorporem modelos de regressão para a combinação de previsões. Assim, o objetivo principal desta dissertação é combinar três previsões individuais (redes neurais, modelos ARIMA e modelos de alisamento exponencial) via média simples, variância mínima e modelos de regressão, comparando as três previsões combinadas com suas previsões individuais. Estas comparações serão avaliadas em duas situações: em séries simuladas (estacionárias) e em uma série de dados reais (não estacionária) de uma empresa que realiza auditorias médicas. As medidas empregadas para a escolha do método mais preciso são MAE, MAPE, RMSE e o coeficiente U de Theil. Os resultados obtidos enfatizam a melhoria das previsões quando estas são combinadas por regressão, tanto para séries convergentes quanto para a série divergente. / Forecasting is a key tool for ensuring the efficiency of management in organizations, directly affecting business profitability. The current competitive corporative level requires increasingly accurate predictions. In this context, the combination of forecasts has improved forecast accuracy. Through a literature review on the approaches of combining forecasts, we identified a lack of comparative studies that incorporate regression models for combining forecasts. Thus, the main objective of this dissertation is to combine three individual forecasts (neural networks, ARIMA models and exponential smoothing models) via simple average, minimum variance and regression models, comparing the three combined forecasts with their individual forecasts. These comparisons are evaluated in two situations: in simulated series (converging) and in series of real data (divergent) from a company that performs medical audits. The measures used to identify the best method are MAE, MAPE, RMSE and Theil’s U coefficient. Results from combined methods improved the predictions in both convergent and divergent series.

A previsão de demanda é uma das principais ferramentas para a eficiência do gerenciamento das organizações, afetando diretamente a lucratividade do negócio. O atual nível competitivo das empresas requer previsões cada vez mais acuradas, sedo estas um diferencial para o sucesso empresarial. Neste contexto, a combinação de previsões se tornou um dos principais métodos empregados no intuito de melhorar a precisão das previsões. Através de uma revisão da literatura sobre as abordagens da combinação de previsões, identificou-se uma carência de estudos comparativos que incorporem modelos de regressão para a combinação de previsões. Assim, o objetivo principal desta dissertação é combinar três previsões individuais (redes neurais, modelos ARIMA e modelos de alisamento exponencial) via média simples, variância mínima e modelos de regressão, comparando as três previsões combinadas com suas previsões individuais. Estas comparações serão avaliadas em duas situações: em séries simuladas (estacionárias) e em uma série de dados reais (não estacionária) de uma empresa que realiza auditorias médicas. As medidas empregadas para a escolha do método mais preciso são MAE, MAPE, RMSE e o coeficiente U de Theil. Os resultados obtidos enfatizam a melhoria das previsões quando estas são combinadas por regressão, tanto para séries convergentes quanto para a série divergente. / Forecasting is a key tool for ensuring the efficiency of management in organizations, directly affecting business profitability. The current competitive corporative level requires increasingly accurate predictions. In this context, the combination of forecasts has improved forecast accuracy. Through a literature review on the approaches of combining forecasts, we identified a lack of comparative studies that incorporate regression models for combining forecasts. Thus, the main objective of this dissertation is to combine three individual forecasts (neural networks, ARIMA models and exponential smoothing models) via simple average, minimum variance and regression models, comparing the three combined forecasts with their individual forecasts. These comparisons are evaluated in two situations: in simulated series (converging) and in series of real data (divergent) from a company that performs medical audits. The measures used to identify the best method are MAE, MAPE, RMSE and Theil’s U coefficient. Results from combined methods improved the predictions in both convergent and divergent series.

Au cours des années récentes, les véhicules électriques sont revenus sur le devant de la scène des politiques publiques en matière de transport. Considérés comme un remède possible à diverses préoccupations pressantes des pouvoirs publics, ils bénéficient d'un soutien croissant de leur part. De telles mesures de soutien demeurent contestées : en effet, leur impact sur le décollage effectif des ventes, leur soutenabilité, leur utilité et leur justification sont loin d'aller de soi. Cette étude vise à éclairer l'impact des politiques publiques destinées à influencer la demande sur i) le taux de pénétration des véhicules électriques auprès des ménages français, et ii) les finances publiques. Dans un premier temps sera brossé le tableau du contexte dans lequel les véhicules électriques ont vocation à se développer. Il sera proposé un panorama large des opportunités potentielles offertes par l'introduction des véhicules électriques. Une revue internationale des politiques publiques est conduite, qui décrit les leviers de politique publique qui sont aujourd'hui actionnés en soutien au véhicule électrique de par le monde. L'accent y est mis sur les mesures destinées à agir sur la demande. Des conclusions préliminaires seront proposées sur l'efficacité de ces mesures au regard des taux observés de pénétration du véhicule électrique. Dans un deuxième temps, l'étude s'attache à évaluer le marché potentiel des véhicules électriques auprès des ménages français. L'analyse porte non seulement sur les déterminants financiers de la demande, mais aussi sur les obstacles socio-économiques à l'adoption des véhicules électriques par ces ménages. S'appuyant sur une analyse par scénarios qui permet de rendre compte des nombreuses incertitudes relatives aux évolutions à prévoir des véhicules, des coûts et des tendances de marché, une prévision du potentiel de demande à l'horizon 2023 est avancée. L'approche désagrégée qui est appliquée à partir de la base de données de l'Enquête Nationale Transports et Déplacements 2007/2008 permet d'identifier les combinaisons de instruments financiers de politique publique les plus à même de garantir certains niveaux de pénétration du véhicule électrique dans la prochaine décennie. Enfin, l'impact sur les finances publiques du remplacement d'un véhicule conventionnel par un véhicule électrique est étudié. L'analyse porte à la fois sur les phases de production et d'usage du véhicule. Le modèle d'évaluation développé à cet effet tient compte des impacts directs et indirects sur les finances publiques. Sont pris en compte les subventions directes à l'achat, les allègements fiscaux, les recettes fiscales, ainsi que les effets sur l'emploi. Les conclusions et observations tirées de l'étude permettent de formuler diverses suggestions à l'attention des constructeurs automobiles et des décideurs publics affichant la volonté de soutenir l'essor du véhicule électrique / In recent years, electric vehicles have come to the forefront of public transport policies. They are seen as remedy for various pressing public concerns and are thus increasingly benefiting from supportive policy measures. Such measures remain contested: their impact on actual vehicle uptake rates, their sustainability, usefulness and justification are far from being self-evident. This study aims at uncovering the effect of financial demand-side public policy measures on i) the uptake rate of electric vehicles among private households in France, and ii) the public budget. First, the context within which electric vehicles are to evolve is sketched. A comprehensive overview of the potential opportunities that come with the introduction of electric vehicles is given. An international policy review depicts public policy levers that are currently deployed in order to support the uptake of electric vehicles. A focus is put on financial demand-side measures. Preliminary conclusions on their effectiveness with regards to observed electric vehicle uptake rates in the various countries reviewed are drawn. Next, the potential market for electric vehicles among French households is explored. Besides financial aspects, socio-economic obstacles to electric vehicle uptake among private households are analysed. With the aid of scenario analysis that accounts for the many uncertainties with regards to future vehicle developments, costs and market trends, a forecast of the electric vehicles' potential up until 2023 is given. The applied disaggregate approach based on the database of the French National Transport Survey 2007/2008 allows identifying the most promising sets of financial public policy measures that are likely to guarantee certain electric vehicle uptake rates over the next decade. Lastly, the effect of replacing one conventional vehicle by one electric vehicle on the public budget is investigated. Both, vehicle manufacture and use aspects are considered. The set up valuation model hereby accounts for direct and indirect financial impacts on the public budget. These comprise direct purchase subsidies, tax breaks, and tax income, as well as effects of changing employment situations that alter the amount of social contributions and unemployment benefits .The study's findings and considerations allow for various suggestions for vehicle manufacturers and policy makers willing to support the uptake of electric vehicles. These are listed in the conclusions section which also sketches directions for further research

<div>As on 2017, US Energy Information Administration (US EIA) claims that 50 % of the total US energy consumption are contributed by Commercial and Industrial (C&I) end-users.</div><div>Most of the energy consumption by these users is in the form of the electric power. Electric utilities, who usually supply the electric power, tend to care about the power consumption profiles of these users mainly because of the scale of consumption and their significant contribution</div><div>towards the system peak. Predicting and managing the peaks of C&I users is crucial both for the users themselves and for utility companies.</div><div>In this research, we aim to understand and predict the daily peaks of individual C&I users. To empirically understand the statistical characteristics of the peaks, we perform an extensive exploratory data analysis using a real power consumption time series dataset. To accurately predict the peaks, we investigate indirect and direct learning approaches. In the indirect approach, daily peaks are identified after forecasting the entire time series for the day whereas in the direct approach, the daily peaks are directly predicted based on the historical data available for different users during different days of the week. The machine learning models used in this research are based on Simple Linear Regression (SLR), Multiple Linear Regression (MLR), and Artificial Neural Networks (ANN).</div>

The present project of applied research has as main problem the synchronization of information in the logistics chain of the company Ceramic Lima, the areas involved in this project are purchases, planning, production and storage, since the processes as a whole did not achieve the desired objective. From the manufacturing point of view, this project has logistic positions that will allow opting for critical analysis, since it shows diverse behaviors in the supply chain. In this way, the fluidity of information between each operating unit is of vital importance, adding the efforts of human capital, financial and natural resources. The motivation to carry out this research was given in order to demonstrate the theoretical knowledge learned throughout the academic training providing technical, operational and economic solutions. In this sense, this article presents the solutions for each eventuality that occurred along the logistics chain; evaluating the viability through an external and internal economic impact. / El presente proyecto de investigación aplicada tiene como problema principal la sincronización de información en la cadena logística de la empresa Cerámica Lima, las áreas involucradas en este proyecto son compras, planeamiento, producción y almacén, ya que los procesos en conjunto no lograban el objetivo deseado. Desde el punto de vista manufacturero este proyecto tiene posturas logísticas que permitirá optar por análisis críticos, ya que muestra diversos comportamientos en la cadena de suministro. De esta manera, la fluidez de información entre cada unidad operativa es de vital importancia, sumándose los esfuerzos de capital humano, recursos financieros y naturales La motivación de llevar a cabo esta investigación se dio con el fin de demostrar los conocimientos teóricos aprendidos a lo largo de la formación académica aportando soluciones técnicas, operativas y económicas. En este sentido, este artículo presenta las soluciones ante cada eventualidad ocurrida a lo largo de la cadena logística; evaluando la viabilidad a través de un impacto económico tanto externo e interno. / Tesis

Orientador: Sergio Valdir Bajay / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-06T05:07:08Z (GMT). No. of bitstreams: 1 Carvalho_ClaudioBezerrade_D.pdf: 1717945 bytes, checksum: 278bbb29104ec96389a1e6616a1dc209 (MD5) Previous issue date: 2005 / Resumo: Este trabalho traz uma avaliação critica do planejamento energético de longo prazo realizado no Pais nos últimos anos e aponta tanto para a necessidade de uma melhor integração entre as atividades de planejamento energético, elaboração de políticas públicas e regulação dos mercados de energia, como para uma premente descentralização do processo de planejamento. Como resultados desta análise e com base em experiências bem sucedidas no exterior, são propostos avanços metodológicos para a elaboração de futuras projeções e o desenvolvimento de um modelo integrado de projeção da demanda e da oferta de energéticos. Como a aplicação de tal modelo está vinculada à utilização de uma base de dados ampla e consistente, é proposto o desenvolvimento de um sistema nacional de informações energéticas, integrado a um sistema de informações executivas, cujo objetivo é servir de suporte para as atividades desenvolvidas pelo Ministério de Minas e Energia. Discute-se os vários métodos de tratamento das incertezas nos modelos energéticos, com destaques para a elaboração de cenários alternativos de desenvolvimento e para o uso da técnica Delphi de levantamento de opiniões de especialistas. Monta-se, por fim, à guisa de um estudo de caso que visa contribuir para o necessário processo de descentralização do planejamento energético no País, cenários alternativos de desenvolvimento para a projeção da demanda energética do Estado da Bahia, de uma forma concatenada com cenários semelhantes no âmbito nacional / Abstract: This work brings a critical evaluation of the long-term energy planning carried out in the country in the last years, pointing out both for the need of a better integration of the activities concerning energy planning, policy making and regulation of energy markets, and for an urgent decentralization of the planning process. As results of this analysis and based on successful experiences abroad,methodological advances are proposed for the elaboration of future forecasts, together with the development of an integrated model for forecasting energy demand and supply. As the application of such a model requires a broad and consistent data basis, setting up a national system of energy information is proposed, integrated to a system of executive information, aimed to support the activities of the Ministry of Mines and Energy. The several methods for treating uncertainties in energy modeling are discussed, with emphasis on the elaboration of alternative development scenarios and the use of the Delphi technique for collecting and processing the opinions of specialists. At the end, alternative development scenarios for forecasting the energy demand in the State of Bahia, linked to similar scenarios at the national level, are elaborated, as a study case aimed to contribute for the necessary decentralization process of energy planning in the country / Doutorado / Planejamento de Sistemas Energeticos / Doutor em Planejamento de Sistemas Energéticos

Accurate demand forecasting is a key capability for a manufacturing organization, more so, a semiconductor manufacturer. Many crucial decisions are based on demand forecasts. The semiconductor industry is characterized by very short product lifecycles (10 to 24 months) and extremely uncertain demand. The pace at which both the manufacturing technology and the product design changes, induce change in manufacturing throughput and potential demand. Well known methods like exponential smoothing, moving average, weighted moving average, ARMA, ARIMA, econometric methods and neural networks have been used in industry with varying degrees of success. We propose a novel forecasting technique which is based on Support Vector Regression (SVR). Specifically, we formulate ν-SVR models for semiconductor product demand data. We propose a 3-phased input vector modeling approach to comprehend demand characteristics learnt while building a standard ARIMA model on the data. Forecasting Experimentations are done for different semiconductor product demand data like 32 & 64 bit CPU products, 32bit Micro controller units, DSP for cellular products, NAND and NOR Flash Products. Demand data was provided by SRC(Semiconductor Research Consortium) Member Companies. Demand data was actual sales recorded at every month. Model performance is judged based on different performance metrics used in extant literature. Results of experimentation show that compared to other demand forecasting techniques ν-SVR can significantly reduce both mean absolute percentage errors and normalized mean-squared errors of forecasts. ν-SVR with our 3-phased input vector modeling approach performs better than standard ARIMA and simple ν-SVR models in most of the cases.