Prediksi Kebutuhan Darah Menggunakan Metode ARIMA Dengan Mempertimbangkan Faktor Deterioration

Abstract

Blood has various uses that are very important for the human body. However, blood can only be donated in limited quantities because it can only be produced by humans. This blood donation activity is organized by the Blood Transfusion Unit (UTD), one of which is UTD PMI Surabaya. UTD PMI Surabaya also produces and stores blood products and distributes them to hospitals or directly to patients who need blood. Due to the uncertain amount of blood demand, UTD PMI Surabaya needs to make predictions in order to meet the blood needs. But blood has an expiry that also needs to be considered. Meanwhile, the blood demand prediction system still uses human estimates. Therefore, an information system is needed that can assist in predicting the need for blood by considering the expiration date of the blood. The application made in this research is a web application that can assist in making predictions using the ARIMA method. Then a calculation will be carried out by considering the deterioration of the blood to determine the need for blood in the next month. Based on the test results, the application is able to predict blood needs. The ARIMA model used to predict WB blood components is ARIMA (7,0,6) with a RMSE of 58,91. While the ARIMA model used to predict the TC blood component is ARIMA (5,0,6) with a RMSE of 272,46.