Análise preditiva de despesas com o Programa de Participação nos Resultados (PPR) a partir de Demonstrativo do Resultado do Exercício (DRE) utilizando modelos de Aprendizado de Máquina (AM)

Abstract

Companies face significant challenges in managing expenses from bonuses. Therefore, this study aims to predict the expenses of the Profit Sharing Program (PPR), using financial data from the Income Statement (DRE) of a small company. The techniques adopted, such as SARIMAX, Prophet and Random Forest, were evaluated in terms of accuracy, revealing potential improvements in expense control and operational sustainability. This work contributes to the strategic use of Machine Learning (ML) in business scenarios.