APPLICATION OF MACHINE LEARNING TECHNIQUES TO MITIGATE UNCERTAINTY IN THE HEALTH AND MEDICAL SUPPLY CHAIN

Despite ongoing efforts by the World Health Organization, managing blood transfusion and distribution remains a critical challenge, especially during periods of high demand and limited inventory. Minimizing uncertainty around blood demand, wastage, and shortages is therefore a primary objective. This study proposes an intelligent, platform-based approach to develop a resilient blood supply chain by forecasting both collection volumes and demand levels. It further aims to reduce blood wastage and shortages by balancing collection and distribution through effective inventory management. Leveraging machine learning algorithms and time series forecasting models, the AI-driven decision support system provides accurate, reliable predictions and identifies key clinical predictors essential for forecasting blood demand. Compared to historical data from the previous year, the integrated system achieved an 11% increase in blood collection, a 20% reduction in inventory wastage, and a decreased incidence of shortages.