Engineering Student CGPA group classification by Machine Learning

Abstract

Students’ performance evaluation turns out to be seriously difficult. The purpose of this paper is to explore students’ characteristics that are related to the student's CGPA (Cumulative Grade Point Average) groups [CGPA was classified into 3 CGPA groups S: Scholarship and PSU (Public Sector Undertakings) eligibility (CGPA>=8), P: PSU eligibility(8>CGPA>=6), N: not eligible for PSU(CGPA<6)] .Publicly available open access student characteristics like gender, lateral entry, branch, if college is government or private were taken into account as possible determinants of CGPA groups. A Real-World data set of 5102 engineering students who completed their undergraduate, spread across 15 colleges in Guru Gobind Singh Indraprastha University were analyzed. Various classifier algorithms were used to find the most accurate model. The insights of this paper have some significant implications. The finding revealed that in some branches around half of the students are in the Scholarship group, others have below a fifth of the total students, Even gender has disparity around half female students are in the Scholarship group but only about a fourth of boys are in the latter group. This investigation adds to the literature on undergraduate engineering students' academic performance. The discoveries of this examination might be valuable for fine-tuning CGPA criteria by characteristics, college administration, and future analysts. This can also help to select more students for the scholarship who may not satisfy the scholarship criteria.