Development of a predictive model to estimate disbursements and unit costs of sale in a financial institution using Machine Learning

The accurate estimation of sales and their respective costs is essential for strategic decision-making in companies, so optimizing its calculation through advanced techniques is key to improving financial management. This research applies machine learning algorithms in the prediction of sales of financial products and unit costs of sale, using classification and regression models based on Random Forest. Through data preprocessing and hyperparameter optimization with GridSearchCV, a significant improvement in the accuracy of predictions is achieved, reducing time and effort compared to traditional methods.