Modelagem preditiva para identificação da probabilidade de reativação de revendedores em multinacional do setor cosmético

Abstract

Door to door salesmen are one of the most efficient ways that cosmetic companies have found to spread their products, however, finding, retaining and reactivating salespeople só that they continue to sell is still a challenge. Throughout this work, machine learning models were built in Python language, based on three types of algorithms, in order to be able to predict with greater certainty who are the representatives who really have a chance to continue reselling cosmetic products even after some time without new orders.