Customer Segmentation and Business Sales Forecasting using Machine Learning for Business Development

Abstract

This study explores the application of machine learning techniques for business development, focusing on sales prediction and customer segmentation, using a Walmart dataset. Performance metrics include Mean Absolute Error (MAE) and R2 scores. Our hybrid approach combines the BIRCH algorithm with time-lagged machine learning (TL-ML). The results reveal that customer segmentation significantly improves model performance across all metrics. Among the techniques tested, models incorporating customer segmentation (CS-RFR and CS-TL-ML) outperform standard Random Forest Regressor models. Specifically, CS-TL-ML shows a slight advantage in terms of both lower MAE and higher R2 scores, confirming its efficacy for sales prediction and customer segmentation tasks.