Abstract: Demand forecasting is essential for supply chain management and inventory control, but complex patterns and big datasets are difficult for standard statistical approaches to handle. In order to improve prediction accuracy, this study suggests a machine learning-based strategy that makes use of models like Random Forest Regressor, Gradient Boosting Regressor, Support Vector Regression, and Neural Networks. Preprocessing of the dataset includes train-test separation, scaling, label encoding, and data cleaning. Model performance is evaluated using evaluation measures like accuracy, precision, and recall, while demand patterns are revealed using visualisations like heatmaps and histograms. Neural networks and ensemble learning are combined to enhance predictions even further.This method offers a scalable and dependable demand forecasting solution by bridging the gap between traditional approaches and contemporary machine learning techniques. Businesses may improve inventory management, lessen stock imbalances, and boost profitability through improved resource utilisation and demand prediction by successfully identifying important influencing elements such product classifications and warehouse locations.
Keywords: Motivation of Machine learning Support Vector Regression, Lasso Regression, Random Forest, Gradient Boosting.
Copyright Form