Multi Layer Perceptron Algorithm as a Prediction of Adzkia University Students' Graduation on Time Based on Gender, Semester Achievement Index, and Number of Credits Taken

Abstract

In the college accreditation process, student academic graduation is used as an assessment criterion. Stored student graduation data can be used to make predictions in the future. Predictions can be measured using AI to produce accurate predictions based on student data for six semesters, namely gender, IPS1-6, and SKS1-6. To predict students' graduation status, this research uses a Multi-Layer Perceptron Artificial Neural Network (MLP JST). Used as a predicator, the feature consists of thirteen attributes. A binary value of one indicates timely pass and 0 indicates not timely pass in the target class. The proposed MLP JST consists of three layers: an input layer consisting of thirteen neurons, a hidden layer consisting of twelve neurons, and an output layer consisting of one neuron. The application used through Python Google Colab. A total of 100 pochs were used to provide instructions. The results include weights for each neuron in the MLP. The model accuracy metric value of 95% indicates that the prediction has a good level of accuracy.