APPLICATION OF ANN MODEL WITH BACKPROPAGATION ALGORITHM FOR PREDICTION LATE PAYMENT LOANS AT PNM MEKAAR
Abstract
Study This aiming For overcome challenge the with building an Artificial Neural Network (ANN) model using algorithm Backpropagation For predict delay payment loans at PNM Mekaar branch Simalungun based on variables related or features that have been determined . ANN model with algorithm Backpropagation built? own optimal performance . The model is built with milk layer ( hidden layer1 = 80 with activation RELu and input layer = 7, hidden layer2 = 8 with dropout = 0.5 and activation RELu , hidden layer3 = 8 with dropout = 0.3 and activation RELu , hidden layer4 = 8 with dropout = 0.1 and activation RELu , hidden layer5 = 1 with activation Sigmoid ). The results of the model testing after evaluated with use testing data , obtained score accuracy reached 99.1%, the value precision score 100%, value recall score 98.9%, and f1-score by 99.4%. The model is trained use batch size parameters 64, optimizer RMSprop , learning rate 0.001, and number of epoch 100.
