A busca da correta modelagem e previsão de volatilidade em séries financeiras é o que motiva grande parte dos analistas e gestores de carteiras. Esta dissertação buscou, portanto comparar dois tipos de modelos de volatilidade - determinística e estocástica - para as três principais séries de retornos de ações do setor siderúrgico brasileiro, quais sejam: Gerdau PN (GGBR4), Usiminas PN (USIM5) e CSN ON (CSNA3). Os três ativos apresentaram estruturas semelhantes para suas volatilidades. Para as três séries foram encontradas especificações determinísticas do tipo AR (1) - EGARCH (1,1) e AR (1) - TGARCH (0,1), ambas com volatilidades estimadas muito próximas. No caso estocástico optou-se por um modelo AR (1) - SV Estacionário para as três séries de retornos. A maior persistência foi observada no ativo da Gerdau, mostrando que um choque sobre o ativo da Gerdau demora mais a se dissipar do que um choque de mesma magnitude sobre os ativos de Usiminas e CSN. Quanto ao efeito alavancagem, a ação da Usiminas apresentou o maior resultado estimado, mostrando que retornos negativos em um dado instante t geram maior volatilidade no período seguinte (t+1) sobre o ativo da Usiminas. Por último comparou-se a qualidade preditiva das duas classes de modelos de volatilidade por meio de previsões um passo à frente durante 21 dias utilizando-se três estatísticas de previsão - erro médio (ME), raiz do erro quadrático médio (RMSE) e erro absoluto médio (MAE). Para o ativo USIM5 as três estatísticas sugerem que o modelo escolhido deve ser o estocástico. Para o ativo GGBR4 e CSNA3 o ME sugere que o modelo escolhido deve ser o determinístico e o RMSE e o MAE sugerem que o modelo escolhido deve ser o estocástico. / The accurate modeling and forecasting of volatility in financial series is what motivates most analysts and portfolio managers. This dissertation sought therefore to compare two types of volatility models - deterministic and stochastic - for three major series of stock returns of the Brazilian steel industry, namely: Gerdau PN (GGBR4), Usiminas PN (USIM5) and CSN ON (CSNA3). The three assets had similar structures to their volatilities. For all the series we found deterministic specifications of the type AR (1) - EGARCH (1,1) and AR (1) - TGARCH (0.1), both with very close volatility estimates. In the stochastic case we chose a model AR (1) - SV Stationary for the three sets of returns. The highest persistence was observed in the asset of Gerdau, showing that a shock on this asset takes longer to dissipate than a clash of the same magnitude on the assets of Usiminas and CSN. For the leverage effect, the series of Usiminas had the highest estimated results, showing that negative returns in a given time t generate greater volatility in period (t +1) on the asset of Usiminas. Finally we compared the predictive quality of the two classes of volatility models through a one step ahead forecast for 21 days using three statistics for forecasting - mean error (ME), mean squared error (RMSEA) and mean absolute error (MAE). For the asset USIM5 the three statistics suggest that the chosen model should be the stochastic. For the assets GGBR4 and CSNA3 the ME suggests that the chosen model should be the deterministic and the RMSE and MAE suggest that the chosen model should be the stochastic.

A busca da correta modelagem e previsão de volatilidade em séries financeiras é o que motiva grande parte dos analistas e gestores de carteiras. Esta dissertação buscou, portanto comparar dois tipos de modelos de volatilidade - determinística e estocástica - para as três principais séries de retornos de ações do setor siderúrgico brasileiro, quais sejam: Gerdau PN (GGBR4), Usiminas PN (USIM5) e CSN ON (CSNA3). Os três ativos apresentaram estruturas semelhantes para suas volatilidades. Para as três séries foram encontradas especificações determinísticas do tipo AR (1) - EGARCH (1,1) e AR (1) - TGARCH (0,1), ambas com volatilidades estimadas muito próximas. No caso estocástico optou-se por um modelo AR (1) - SV Estacionário para as três séries de retornos. A maior persistência foi observada no ativo da Gerdau, mostrando que um choque sobre o ativo da Gerdau demora mais a se dissipar do que um choque de mesma magnitude sobre os ativos de Usiminas e CSN. Quanto ao efeito alavancagem, a ação da Usiminas apresentou o maior resultado estimado, mostrando que retornos negativos em um dado instante t geram maior volatilidade no período seguinte (t+1) sobre o ativo da Usiminas. Por último comparou-se a qualidade preditiva das duas classes de modelos de volatilidade por meio de previsões um passo à frente durante 21 dias utilizando-se três estatísticas de previsão - erro médio (ME), raiz do erro quadrático médio (RMSE) e erro absoluto médio (MAE). Para o ativo USIM5 as três estatísticas sugerem que o modelo escolhido deve ser o estocástico. Para o ativo GGBR4 e CSNA3 o ME sugere que o modelo escolhido deve ser o determinístico e o RMSE e o MAE sugerem que o modelo escolhido deve ser o estocástico. / The accurate modeling and forecasting of volatility in financial series is what motivates most analysts and portfolio managers. This dissertation sought therefore to compare two types of volatility models - deterministic and stochastic - for three major series of stock returns of the Brazilian steel industry, namely: Gerdau PN (GGBR4), Usiminas PN (USIM5) and CSN ON (CSNA3). The three assets had similar structures to their volatilities. For all the series we found deterministic specifications of the type AR (1) - EGARCH (1,1) and AR (1) - TGARCH (0.1), both with very close volatility estimates. In the stochastic case we chose a model AR (1) - SV Stationary for the three sets of returns. The highest persistence was observed in the asset of Gerdau, showing that a shock on this asset takes longer to dissipate than a clash of the same magnitude on the assets of Usiminas and CSN. For the leverage effect, the series of Usiminas had the highest estimated results, showing that negative returns in a given time t generate greater volatility in period (t +1) on the asset of Usiminas. Finally we compared the predictive quality of the two classes of volatility models through a one step ahead forecast for 21 days using three statistics for forecasting - mean error (ME), mean squared error (RMSEA) and mean absolute error (MAE). For the asset USIM5 the three statistics suggest that the chosen model should be the stochastic. For the assets GGBR4 and CSNA3 the ME suggests that the chosen model should be the deterministic and the RMSE and MAE suggest that the chosen model should be the stochastic.

Summary Along the development of bussines and market environment, where supply is superior to demand todays world brings also higher standards of a customer. Customers aren't willing to make any concessions or tolerate mistakes of suppliers, and still demand highest quality. Companies are exposed to big competition and searching for possibilities how to prevail on market. One of these possibilities is high quality product and service. The amount of companies, which have quality certificates proves how important is to have good reliable product. The product and service quality became an important rival factor and often evenly matched, or even more important then price. Graduation theses is concerning area of a management quality as a tool for upgrading efficiency of a company and quality management with the purpose to gain and maintain customers satisfaction. The objective of this theses is to evaluate actual situation of quality management and to propose general changes in managing, leading to create quality system which enables acquirement of the quality certificate. To achieve objective of this theses, an analyses of current management in a certain company was elaborated. On the base of this analyses general changes in quality management were proposed by selected elements of the norm ISO 9001:2000. The outcome of this work proves, that there are many steps and changes needed to be accomplished to achieve in the quality management which is true to the norm CSN EN ISO 9001:2001.

A busca da correta modelagem e previsão de volatilidade em séries financeiras é o que motiva grande parte dos analistas e gestores de carteiras. Esta dissertação buscou, portanto comparar dois tipos de modelos de volatilidade - determinística e estocástica - para as três principais séries de retornos de ações do setor siderúrgico brasileiro, quais sejam: Gerdau PN (GGBR4), Usiminas PN (USIM5) e CSN ON (CSNA3). Os três ativos apresentaram estruturas semelhantes para suas volatilidades. Para as três séries foram encontradas especificações determinísticas do tipo AR (1) - EGARCH (1,1) e AR (1) - TGARCH (0,1), ambas com volatilidades estimadas muito próximas. No caso estocástico optou-se por um modelo AR (1) - SV Estacionário para as três séries de retornos. A maior persistência foi observada no ativo da Gerdau, mostrando que um choque sobre o ativo da Gerdau demora mais a se dissipar do que um choque de mesma magnitude sobre os ativos de Usiminas e CSN. Quanto ao efeito alavancagem, a ação da Usiminas apresentou o maior resultado estimado, mostrando que retornos negativos em um dado instante t geram maior volatilidade no período seguinte (t+1) sobre o ativo da Usiminas. Por último comparou-se a qualidade preditiva das duas classes de modelos de volatilidade por meio de previsões um passo à frente durante 21 dias utilizando-se três estatísticas de previsão - erro médio (ME), raiz do erro quadrático médio (RMSE) e erro absoluto médio (MAE). Para o ativo USIM5 as três estatísticas sugerem que o modelo escolhido deve ser o estocástico. Para o ativo GGBR4 e CSNA3 o ME sugere que o modelo escolhido deve ser o determinístico e o RMSE e o MAE sugerem que o modelo escolhido deve ser o estocástico. / The accurate modeling and forecasting of volatility in financial series is what motivates most analysts and portfolio managers. This dissertation sought therefore to compare two types of volatility models - deterministic and stochastic - for three major series of stock returns of the Brazilian steel industry, namely: Gerdau PN (GGBR4), Usiminas PN (USIM5) and CSN ON (CSNA3). The three assets had similar structures to their volatilities. For all the series we found deterministic specifications of the type AR (1) - EGARCH (1,1) and AR (1) - TGARCH (0.1), both with very close volatility estimates. In the stochastic case we chose a model AR (1) - SV Stationary for the three sets of returns. The highest persistence was observed in the asset of Gerdau, showing that a shock on this asset takes longer to dissipate than a clash of the same magnitude on the assets of Usiminas and CSN. For the leverage effect, the series of Usiminas had the highest estimated results, showing that negative returns in a given time t generate greater volatility in period (t +1) on the asset of Usiminas. Finally we compared the predictive quality of the two classes of volatility models through a one step ahead forecast for 21 days using three statistics for forecasting - mean error (ME), mean squared error (RMSEA) and mean absolute error (MAE). For the asset USIM5 the three statistics suggest that the chosen model should be the stochastic. For the assets GGBR4 and CSNA3 the ME suggests that the chosen model should be the deterministic and the RMSE and MAE suggest that the chosen model should be the stochastic.

Supercomputer in Betrieb Chemnitzer Studentennetz eingeweiht Neue Compute-Server Neuer Dienst: PC-Integration TeX -Service des URZ Software-News Advent, Advent - Geschichtenzeit

This diploma thesis is focused on design of adsorber. Rule CSN EN 13 345 is used to calculate of design adsorber. Finite element method (FEM) is used tor stress analysis a categoration of the stress.

Das Chemnitzer StudentenNetz (CSN) verwaltet ein mittelgroßes Netzwerk in den Wohnheimen des Studentenwerkes Chemnitz-Zwickau. Mitarbeitern und Interessenten erhalten einen Überblick und Einblick in die vorhandene Netzstruktur und die Konfiguration der verwendeten Technik. Es wird auch auf mögliche Szenarien für zukünftige Entwicklungen eingegangen.

Aus historischen Gründen war die CSN Datenbank und die darauf zugreifenden Dienste immer auf dem selben Rechner. Zum einen aus Geldmangel, zum anderen, weil die Verteilung der Konfiguration und Zugriffssteuerung zur Datenbank ein ungelöstes Problem ist. Aufgabe dieser Arbeit ist die physikalische und logische Trennung der Firewall (und des Shapers) von der Datenbank. Dazu muss ein Dienst geschaffen werden, der die Konfigurationsinformationen für die Firewall und potentiell andere Applikationen bereitstellt. Der Zugriff auf diese Informationen muss vor Dritten geschützt werden. Im Weiteren soll eine Migrationstrategie entworfen werden, wie der Übergang zu der skizzierten Lösung bewerkstelligt werden kann.

Voice commands are used in today's society to offer services like putting events into a calendar, tell you about the weather and to control the lights at home. This project tries to extend the possibilities of voice commands by improving an earlier proof of concept system that interprets intention given in natural language to program code. This improvement was made by mixing linguistic methods and neural networks to increase accuracy and flexibility of the interpretation of input. A user testing phase was made to conclude if the improvement would attract users to the interface. The results showed possibilities of educational purposes for computational thinking and the issues to overcome to become a general programming tool.

With society's growing concern of the environmental impact that consumerism has today, consumers are more aware than ever of how their consumption habits affect their surroundings. This study is based on the consumer's journey of choosing products based on their environmental impact and the complications of it. The study will examine a subsection of Greenwashing, called Executional Greenwashing, that is used to subconsciously communicate messages about greenery towards the consumer, inferring their decision making. The goal of the research is to provide an understanding of how consumers react to Fast-Moving Consumer Goods advertisements containing Executional Greenwashing elements and what factors influence their reaction. The study uses primary and secondary data together with conducted interviews to examine how the use of Executional Greenwashing affects consumers' emotions towards advertisements containing these elements. The findings reveal how knowledge and expertise are the main factors to influence how consumers react to Executionally Greenwashed advertisement. The study also provides a conceptual model based on the findings to explain what emotions are affected in consumers. Consumers and researchers can use these findings to gain a further understanding of the effects of Executional Greenwashing and continue upon this research to achieve a complete understanding.