For retail and e-commerce companies to succeed, effective inventory management is crucial. Inadequate demand forecasting frequently leads to issues like supply shortages, overstocking, and higher operating expenses. This study offers a machine learning-based method for forecasting product demand levels in retail inventory systems in order to overcome these difficulties. The suggested approach makes use of past retail data that includes characteristics like product category, units sold, inventory level, pricing details, weather, promotional activities, and seasonal elements. The dataset is analysed using machine learning methods, such as Random Forest and Gradient Boosting, to divide product demand into three groups: Low, Medium, and High. To enhance model performance, data preprocessing methods like feature encoding and normalisation are used. A web-based application that incorporates the trained models enables users to enter product and store data, and receive real-time demand forecasts. According to experimental findings, the Gradient Boosting model outperforms other models in terms of prediction accuracy. The created solution lowers inventory-related risks, increases supply chain efficiency, and assists merchants in making well-informed decisions about stock levels. This study shows how machine learning approaches might improve inventory demand predictions for contemporary retail settings.
Ekambaram et al. (Thu,) studied this question.