Student Performance prediction using Data Mining Techniques

Abstract

Universities must have a way to pick students who will do well in school based on objective criteria used in the admissions process. This study looks at how data mining tools can help colleges decide who to let in by making it easier to predict how well students will do once they are there. From 2016 to 2019, 2,039 students at a Saudi state university's Computer Science and Information College were used to test the new method. outcomes show that prospective students' early success in college can be predicted by looking at data collected before they are admitted like average grade points percentage etc. outcomes also show that a student's score on the probabilistic test is the best predictor of how well they will do in the future. Because of this, this score needs to be given more weight during the selection process.Sometimes the ANN method was more accurate (by more than 79%) than the other ways we looked at to group things.