Paper Contents
Abstract
: Accurate sales forecasting is crucial for retail businesses to optimize inventory management, improve decision-making, and enhance revenue generation. Traditional statistical methods often fail to capture complex sales patterns influenced by multiple factors. This study introduces a machine learning-based sales prediction framework leveraging regression techniques for enhanced forecasting accuracy. The framework employs Decision Tree Regression, Random Forest Regression, and Linear Regression to analyse sales trends across multiple Big Mart stores. The proposed solution achieved a 95.81% accuracy using Random Forest Regressor, outperforming traditional models by 12%. Feature engineering techniques, including label encoding, one-hot encoding, and data normalization, improved data preprocessing efficiency. The sales prediction framework also reduced forecasting errors by 15%, leading to better demand estimation and inventory optimization. Additionally, correlation analysis identified Item MRP, Outlet Type, and Item Visibility as key sales drivers. These results highlight the potential of machine learning in retail analytics, providing a scalable and data-driven approach for improving sales predictions, minimizing losses, and enhancing operational efficiency in real-world retail environments.
Copyright
Copyright © 2025 Vummadi Sai Bhavna . This is an open access article distributed under the Creative Commons Attribution License.
