|
Acadêmico(a): Ricardo Kremer |
Título: Sistema de Apoio à Decisão para Previsões Genéricas Utilizando Técnicas de Data Mining |
|
Abstract: |
The main purpose of this work is to help decision making through a Decision Support System using Data Mining techniques, specifically to do generic forecasts. To aid in this task, a prototype was implemented that allows the user to define a forecast model, where the same can be trained to answer the forecast variables with certain flexibility. For the elaboration of the prototype, the characteristics of Information Systems were analyzed, as well as the Data Warehouses and Data Mining techniques. In the area of Artificial Intelligence, the Neural Networks were also studied, once they are part of Data Mining applied to the forecast. As a consequence of the development of this work, it was confirmed that the use of data Mining along the stages of KDD was very eficient. Tests were conducted and it was possible to develop models of forecast and the use of Neural Networks was put to practice, which proved eficient enough for the process of forecast especially for its great capacitance of generalization. |
|
|