Modelagem e análise da satisfação dos consumidores de uma instituição financeira utilizando aprendizado de máquina
With the market becoming more and more competitive, it becomes necessary to maintain a good relationship with your consumers, a factor that can generate optimization of resources and an immense competitive advantage. Among the markets that depend most on this relationship to keep their customers, make them generate value through word-of-mouth marketing and avoid evasion to competitors are banks and insurance companies. Thus, this article presents a Data Science methodology with the objective of creating a model that can predict the satisfaction of a financial institution’s customers, through its characteristics and behavior described in a database provided by the bank. In a way that allows the company to act with countermeasures to reverse the cases of dissatisfaction. As a proposed solution, three machine learning models were used (Decision Tree, Random Forest and XGBoost) that had a maximum performance of 82.02%, 81.61% and 83.39%, a considerable and efficient performance due the nature of the problem and database.