Predicting Student Academic Performance at Higher Education Using Data Mining: A Systematic Review

Table 3

Summary of the related studies based on predicting student's achievement at graduation time.

Ref	Techniques	Results	Study sample size	Findings
[5]	DT (J48)	Accuracy (69.3%)	339	The CGPA of the first year and three courses of the first year: Introductory math, computer skills, and communication skills are the most influence factors on the graduation CGPA.
[47]	DT and NB	Accuracy (73.41), AUC (66.4%).	79	The findings showed that the NB outperformed DT in predicting the graduation CGPA.
[6]	NB and Hoeffding Tree	Accuracy (91%)	530	Four courses have a significant influence on the CGPA: operating systems, statistics, general physics, computer programming, and algorithms course.
[48]	LR	Accuracy (89.15%)	1841	Third year GPA is the highest influencing feature on the final year graduation GPA.
[49]	NB	Accuracy (43.18%)	2281	Preadmission requirement and the personal student information influence the graduation GPA
[50]	DT	Accuracy (80%)	100	Second year course grade is the highest influencing feature on the final graduation year GPA.
[51]	LR	Correlation coefficient (64%),	957	Preuniversity exams such as SAAT and GAT do not influence the student’s GPA, whereas the high school GPA affects the student’s GPA.
		MAE (0.17)