Abstract:
The long-term sustainability of Higher Education Institutes (HEIs) requires predicting
student At-Risk to provide them with timely intervention and support. Machine Learn ing (ML) techniques can be helpful if we predict their academic performance and identify
At-Risk students. In this research, binary classification is performed on a student dataset
of four years (2017-2021) of three different disciplines: Bachelor in Computer Science
(BSCS), Bachelor in Electrical Engineering (BEEE), and Bachelor in Software Engineer ing (BESE) obtained from SEECS, in National University of Sciences and Technology
(NUST), Islamabad, Pakistan. The research focuses on classifying whether the student
is At-Risk or not at risk in a semester. The student with a CGPA below 2.5 is con sidered At-Risk. ML algorithms: Support Vector Machine (SVM), Logistic Regression,
XgBoost, are used to achieve the target objective to predict At-Risk students. Differ ent factors were involved in affecting the student’s performance such as demographics,
pre-university data, and semester-wise data. . . .
