PREDIKSI KELULUSAN MAHASISWA TEKNIK INFORMATIKA UNIVERSITAS BANTEN JAYA MENGGUNAKAN ALGORITMA NEURAL NETWORK

Abstract

The university strives to provide relevant knowledge. One way the government can use it is to measure the quality of the institution by the number of graduates. The higher the pass rate, the higher the quality of training, which can have a positive impact on the certifications awarded by BAN-PT. This allows researchers to see how research is being conducted at the University of Banten Jaya. To predict graduation rates, students can use a type of artificial neural network algorithm commonly known as neural networks. Artificial neural networks are machine learning techniques developed from Multilayer Perceptron (MLP) and designed to process two-dimensional data. Neural network algorithms belong to the type of deep neural network imaging used. There are several types of neural network techniques. That is, the steps of forward and reverse propagation training. Neural networks are similar to MLPs, but in neural networks each neuron is represented in two dimensions, as opposed to MLP, where each neuron has only one dimension. The results of student graduation in a timely manner and is expected to provide information and can provide input to universities in formulating policies for future improvements.