<![CDATA[Background: Blood availability remains a major challenge in healthcare systems, particularly in developing countries where the demand for blood often exceeds the available supply. Accurate forecasting of blood collection is therefore important to support effective blood inventory management at blood transfusion units. Aims: This study aims to apply the Average-Based Fuzzy Time Series method to forecast the number of collected blood bags at the Blood Transfusion Unit (UTD) of the Indonesian Red Cross in Banda Aceh, both in total and by blood type. Method: Monthly blood collection data from January 2016 to September 2020 were analyzed using the Average-Based Fuzzy Time Series model. The forecasting procedure involved constructing fuzzy intervals using the average-based approach, forming fuzzy logical relationships, and performing defuzzification. Model performance was evaluated using Mean Squared Error (MSE) and Average Forecasting Error Rate (AFER). Result: The second-order model provided the best forecasting performance with an AFER value of 13.67% and an accuracy of approximately 86.33%, producing a prediction of 2054 blood bags for October 2020. Forecasting by blood type yielded predictions of 529 (A), 702 (O), 738 (B), and 154 (AB) blood bags. Conclusion: The results indicate that the Average-Based Fuzzy Time Series method is effective for forecasting blood bag availability and can support planning and management of blood supply at blood transfusion units. Furthermore, the proposed approach has potential applications in defense and emergency contexts by supporting medical logistics planning, improving preparedness, and enhancing the resilience of blood supply systems during military operations and disaster response.]]>
Copyrights © 2026
