Jaringan Syaraf Tiruan Untuk Menentukan Jurusan Di Sekolah Menengah Atas (SMA) (Studi Kasus : SMA Negeri 1 Sungai Penuh)




Students in the Education Unit High School (SMA) must determine when the department will increase from class X to class XI, as noted in the Director General SK Mendikdasmen No.12/C/Kep/TU/2008 dated 12 February 2008 on the Consult Report Preparation learning (LHB) of Students SMA (SBC). Determination of majors students from nine criteria specified. The algorithm used is the Backpropagation with Momentum. Neural network used has an input layer consisting of nine pieces of nodes, one hidden layer with n nodes are randomly determined by the activation function, and an output layer with 2 pieces of input nodes majoring in science and social studies majors with purelin activation function. Data is separated into two parts, ie the data to be trained and the data that will be tested. Training and testing was conducted using the software Matlab 6.1 with the tested network architecture that is 9-2-1, 9-4-1, 9-6-1, and 9-9-1, where the results of processing with matlab menyimpulkam sharing of data 120 for training and 48 data for testing the architectural pattern of 9-9-1 is the best pattern to the percentage of 100% truth.
Key words :  artificial neural networks, backpropagation, schools, and schools