Previsão da demanda por aulas práticas em Centros de Formação de Condutores: aplicação de modelos quantitativos de séries temporais

Abstract

This Final Paper aims to identify the most efficient method for forecasting the demand for practical driving lessons in categories A, B, and D, in two Driving School Centers (CFCs) located in Governador Valadares – MG. The research uses a quantitative approach, applying methods based on time series. Data were extracted from the internal system CFC PRODUTIVO, covering the period from 2018 to 2024, and only effectively conducted classes were considered, as they represent the materialized demand. The applied methods include Simple Moving Average, Weighted Moving Average, Simple Exponential Smoothing, Holt, Holt-Winters Additive, and Holt-Winters Multiplicative. The comparison between the models was based on the Mean Squared Error (MSE) and the Root Mean Squared Error (RMSE). The results allowed for the identification of the most suitable model for each category and unit, contributing to strategic decisions related to productive capacity, operational planning, and the management of service offerings at the CFCs. The detailed analysis demonstrated the importance of accurate demand forecasting to optimize resource allocation and reduce waste, providing valuable support for the business's strategic management.