Human blood is liquid in human body, which functions to transport oxigen needed by cells to the whole body. Considering the important blood function, the Indonesian Red Cross (PMI) has to maintain its blood stock stability to ensure the blood availibility. But the problem that PMI has to encounter with is its blood over-supply which leads to blood disposal. To minimize its unnessary blood disposal, estimation of blood need is required. Data of blood demand is normalized first, then estimation is made using Neural Network Backpropagation. In this study the estimation is made to the blood type of Packet Red Cells (PRC), the blood cells stocked at PMI Kota Surabaya. The best simulation result is at epoch 3000 with function Y = 4542,33 – 1,64595 x – 0,244018 x^2 and an error of 0,020314.
Copyrights © 2018