Big Mart Product Sales Estimator Using Machine Learning

The retail industry significantly benefits from sales prediction since it facilitates the management of inventories and the making of business decisions. The use of machine learning techniques to create a Big Mart Product Sales Estimator for the purpose of predicting product sales at various retail outlets is the central theme of this paper. The model's foundation is built on historical sales data, which includes features of key products and outlets such as product category, weight, visibility, maximum retail price, outlet size, and location. To boost the accuracy of predictions, data preprocessing methods are applied such as handling of missing values, categorical encoding, and feature scaling. Three machine learning models—Linear Regression, Decision Tree, and Random Forest—are created and assessed with baseline performance metrics. The experiments suggest that Random Forest Regression is the best model since it provides higher accuracy and lower prediction error than the other models. The system that has been proposed helps retailers to make informed decisions based on data, to optimize stock levels and to enhance their sales performance overall.