Implementation of The Backpropagation Method and The Kohonen Network to Predict Blood Availability: Case Study in PMI Kota Magelang

The availability of blood stocks at the Indonesian Red Cross or Palang Merah Indonesia (PMI) is a must and absolute for institutions that organize procurement distribution medical purposes. problem stock in PMI Magelang City Branch not ideally available each type, especially type AB, which recent years has been very minimal difficult to obtain. purpose this study was predict AB with software based on artificial neural networks backpropagation Coherent tissues. Artificial Neural Network (JST) used supply AB. Meanwhile, network divide input patterns into groups. application structure consisting 2 neurons, 10 neurons hidden layer, 1 neuron output layer. total amount data 3 (2015-2017), are training data, year testing data. engine predicts using maximum iteration 1,000 epochs, an expletive constant 0.5, momentum 0.9, minimum error rate 0.001. With variations value component, prediction less than 140 bags per generated. resulting weight predicted by method produces production month. Based results against test percentage system accuracy 100%. reduction learning constants addition systems may affect making predictions.