Aplicação de um modelo preditivo de mineração de dados no apoio à decisão de crédito

• Ano de defesa: 2006
• Autor(a) principal: Roberto de Souza Santos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Escoragem de crédito; Ciência da informação; Mineração de dados; Fontes de informação; Controle de crédito; Mineração de dados (Computação)
• Link de acesso: http://hdl.handle.net/1843/VALA-6T7R6V

The present study searches the understanding of the contribution of the knowledge database discovery on supporting the credit decision. For this purpose, its general objective is to create a predictive model capable of identifying information through data mining, as well as to explore how the model can be used. The studys specific objectives are to explore how to plan and execute each one of the tasks required to create the predictive model and to implement it using a data mining methodology. This study also seeks to understand the role of this model on a broader knowledge discovery process over the credit risk at financial institutions. A knowledge process was recognized in which the historical data represents organizational memory, the data mining represents the learning process and the knowledge extracted is the new domain knowledge. The use of data mining, which focuses on the extraction andapplication processes of knowledge database discovery, can support the credit decision when guided by the credit scoring definitions.