Enhancing Demand Forecasting in Retail Supply Chains: A Machine Learning Regression Approach
This investigation discusses the importance of supply chain management and the role of demand forecasting in the business circle and presents a review of literature on demand forecasting techniques, emphasizing the shift from traditional methods to more sophisticated statistical and machine learning approaches. The study aims to contribute to existing knowledge on demand forecasting by utilizing machine learning regressors to predict orders in a Brazilian logistics company. It showed the use of the PyCaret Python library to develop robust regression models and validate key contributing features through feature importance plots. The performance of eighteen models, including Ridge, LASSO, XGBoost, Bayesian Ridge, Linear Regression, Gradient Boosting, KNN, Random Forest, among others, is evaluated using the Mean Absolute Error (MAE) metric.
