Publicação

Predictive Modelling of Merchandising Sales in a Football Club

Ver documento

Detalhes bibliográficos
Resumo:This project develops a weekly merchandising sales forecasting model for a professional football club with the goals of maximizing order quantity from suppliers, avoiding stockouts, and reducing overstock. Based on the CRISP-DM process, historical sales, performance of the club, and weather information were collected, cleaned, and analysed. The extracted variables were used to depict product launches and promotions,so that the last dataset could be weekly aggregated and employed to train and compare the various forecasting models: SARIMAX, XGBoost, LightGBM, and Random Forest. XGBoost performed better than the other models, exhibiting the following performance, RMSE of 84.221, MAE of 46010.88 and an adjusted R² of 0.846, being superior in detecting non-linear relationships and intricate patterns in the data. This study demonstrates how machine learning methodology can become a major value driver of operational efficiency, enabling inventory management and creation of more strategic marketing campaigns, in addition to maximizing fan experience through access to most desirable products.
Autores principais:Carneiro, Marta Francisco da Cunha Mendes
Assunto:Machine Learning Forecasting Demand Time series ARIMA XGBoost Random Forest LightGBM SDG 8 - Decent work and economic growth SDG 9 - Industry, innovation and infrastructure SDG 12 - Responsible production and consumption SDG 17 - Partnerships for the goals
Ano:2025
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
Descrição
Resumo:This project develops a weekly merchandising sales forecasting model for a professional football club with the goals of maximizing order quantity from suppliers, avoiding stockouts, and reducing overstock. Based on the CRISP-DM process, historical sales, performance of the club, and weather information were collected, cleaned, and analysed. The extracted variables were used to depict product launches and promotions,so that the last dataset could be weekly aggregated and employed to train and compare the various forecasting models: SARIMAX, XGBoost, LightGBM, and Random Forest. XGBoost performed better than the other models, exhibiting the following performance, RMSE of 84.221, MAE of 46010.88 and an adjusted R² of 0.846, being superior in detecting non-linear relationships and intricate patterns in the data. This study demonstrates how machine learning methodology can become a major value driver of operational efficiency, enabling inventory management and creation of more strategic marketing campaigns, in addition to maximizing fan experience through access to most desirable products.