| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.19 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A previsão constitui um importante ativo para as empresas, uma vez que ajuda no seu processo operacional e estratégico. A previsão do futuro através de metodologias quantitativas caracteriza-se por ser bastante útil para a antecipação de tomadas de decisão, o que poderá levar a vantagem competitiva e sucesso no universo empresarial. O desenvolvimento das áreas de business intelligence e business analytics têm um forte impacto na implementação destes métodos, uma vez que apresentam importantes ferramentas que se mostram eficientes para a análise e previsão. Este trabalho centra-se em duas análises preditivas, uma referente às vendas totais da empresa “OLI – Sistemas Sanitários, S.A”, e a segunda referente à produção de energia do painel solar da “OLI Moldes, Lda”. No sentido de averiguar métodos preditivos mais eficazes e eficientes para os dois casos de estudo implementados, empregou-se alguns métodos quantitativos de previsão existentes. No primeiro caso, pode-se destacar a implementação do Método de Holt Winters Aditivo, o Método Holt Winters Multiplicativo e o Método ARIMA. No segundo caso foram utilizados métodos causais, como o Modelo de Modelo de Regressão Múltipla, o Modelo de Regressão de Ridge, o Modelo de Regressão de Lasso, o Modelo de Regressão de Elastic Net e o Modelo de Floresta Aleatória. Deste modo, este trabalho conduz a uma elucidação dos conceitos de business analytics e business intelligence, que fortalecem a compreensão dos métodos quantitativos de previsão aplicados.
Forecasting is an important asset for companies, as it helps in their operational and strategic process. Predicting the future through quantitative methodologies is characterized by being very useful for anticipating the decision making, which can lead to competitive advantage and success in the business universe. The development of the business analytics and business intelligence areas has a strong impact on the implementation of these methods, since they present important tools that prove to be efficient for analysis and forecasting. This work focuses on two predictive analyses, one referring to the total sales of the company "OLI - Sistemas Sanitários, S.A", and the second one referring to the energy production of the solar panel of "OLI Moldes, Lda". To ascertain more effective and efficient predictive methods for the two case studies implemented, some existing quantitative forecasting methods were used. In the first case, we can highlight the implementation of the Additive Holt Winters Method, the Multiplicative Holt Winters Method and the ARIMA Method. In the second case, causal methods were used, such as the Multiple Regression Model, the Ridge Regression Model, the Lasso Regression Model, and the Random Forest Model. In this way, this work leads to an elucidation of the concepts of business analytics and business intelligence, which strengthen the understanding of the quantitative forecasting methods applied.
Forecasting is an important asset for companies, as it helps in their operational and strategic process. Predicting the future through quantitative methodologies is characterized by being very useful for anticipating the decision making, which can lead to competitive advantage and success in the business universe. The development of the business analytics and business intelligence areas has a strong impact on the implementation of these methods, since they present important tools that prove to be efficient for analysis and forecasting. This work focuses on two predictive analyses, one referring to the total sales of the company "OLI - Sistemas Sanitários, S.A", and the second one referring to the energy production of the solar panel of "OLI Moldes, Lda". To ascertain more effective and efficient predictive methods for the two case studies implemented, some existing quantitative forecasting methods were used. In the first case, we can highlight the implementation of the Additive Holt Winters Method, the Multiplicative Holt Winters Method and the ARIMA Method. In the second case, causal methods were used, such as the Multiple Regression Model, the Ridge Regression Model, the Lasso Regression Model, and the Random Forest Model. In this way, this work leads to an elucidation of the concepts of business analytics and business intelligence, which strengthen the understanding of the quantitative forecasting methods applied.
Description
Keywords
Business analytics Business intelligence Métodos quantitativos de previsão Quantitative forecasting methods