Application of predictive analytics for identifying customer churn in a company providing payment services in the logistics sector

The thesis focuses on the application of predictive models for customer churn identification in a selected organization. The solution is based on the analysis of available data sources, the creation of appropriate metrics for evaluating customer behavior, and the implementation of machine learning methods. The data undergo a process of validation, adjustment, and preprocessing to ensure their suitability for predictive modeling. Subsequently, predictive models are trained, and their performance is evaluated based on key metrics. The final part evaluates the benefits of the applied approach in business practice and outlines further development of predictive analytics in customer management.