Application of Naïve Bayes Method for Assessing Student Performance

Riza Akhsani Setyo Prayoga (State University of Surabaya, Indonesia)
Fauzan Nusyura (Universiti Kebangsaan Malaysia, Malaysia)
Fiddin Yusfida A’la (Sebelas Maret University, Indonesia)
Mustafa Kamal (Telkom University, Indonesia)
Farhanna Mar'i (State University of Surabaya, Indonesia)

Journal of Advanced in Information and Industrial Technology (JAIIT)
Vol. 8 No. 1: May

Every school has pupils with varying levels of achievement. These differences in achievement can be influenced by several factors, such as the parents’ level of education and the pupils’ readiness for examinations. Furthermore, they can also be influenced by pupils’ abilities in mathematics, writing, and reading. The aim of this study is to classify student performance so that the performance of students at an adequate level or below average can be improved. The method used in this study is Naïve Bayes as a classification method. There are 150 training data points and 50 test data points. Five metrics were evaluated: precision at 94.4%, recall at 94.4%, specificity at 50%, accuracy at 90%, and the F1 score at 94%. This indicates that the model performs well in providing accurate positive predictions. Furthermore, the model is capable of detecting the majority of positive cases effectively.