Table 1 Comparative analysis
References | Variables | Objectives | Level | Dataset | Algorithms | Accuracy | |
|---|---|---|---|---|---|---|---|
Min | Max | ||||||
Asif et al. (2017) | The marks for all the courses that are taught in the four years of the degree programme | Predicting students' performance | Undergraduate students | 210 | DT, 1-NN, NB, NN, RF | NN (62.50%) | NB (83.65%) |
Cruz-Jesus et al. (2020) | Year of the study cycle, gender, age, number of enrolled years in high school, scholarship, internet access, class size, school size, economic level, population density, number of unit courses attended | Predicting students' performance | High schools students | 110627 | ANN, DT, ET, RF, SVM, kNN, LR | LR (81.1%) | SVM (51.2%) |
Fernandes et al. (2019) | Class with persons with special needs, Classroom usage environment, Gender, age (mean), Student benefit, city, neighbourhood, Student with special needs, Grade (mean), Absence (mean) | Predict academic outcomes of student performance | High schools students | Dataset1:19000 Dataset2:19834 | Gradient Boosting Machine | 89.5% | 91.9% |
Hoffait and Schyns (2017) | Gender, Nationality, Studies, Prior schooling, math, scholarship, success | Predicting students at high risk of failure | secondary school students | 2244 | RF, LR, ANN | ANN (70.4%) | RF (90%) |
Rebai et al. (2020) | Socioeconomic status, school type, school location, competition, teacher characteristic (experience, salary), class size, school size, gender, parental education, political context, parental pressure | to identify the key factors that impact schools' academic performance and to explore their relationships | Secondary schools | 105 schools | RT, RF | ||
Ahmad and Shahzadi (2018) | Previous degree marks, Home environment, Study habits Learning skills, Hardworking and Academic interaction | Identification of students in the risk group | Undergraduate students | 300 | MPNN | 95% | |
Musso et al., (2020) | Learning strategies, coping strategies, cognitive factors, social support, background, self-concept, self-satisfaction, use of IT and reading | Grade point average, academic retention, and degree completion | Undergraduate students | 655 | ANN | 60.5% | 80.7% |
Waheed et al., (2020) | Students’ demographics, clickstream events | Pass-fail, withdrawn-pass, distinction-fail, distinction-pass | Undergraduate students | 32593 | ANN, SVM, LR | 84% | 93% |
Xu et al. (2019) | Internet usage behaviours comprise online time, internet connection frequency, internet traffic volume, and online time | Predicting students' performance | Undergraduate students | 4000 | DT, NN, SVM | 71% | 76% |
Bernacki et al. (2020) | Log records in the learning management system | Predict achievement | Undergradeate students | 337 | LR, NB, J-48 DT, J-Rip DT | J-48 (53.71%) | LR (67.36%) |
Burgos et al. (2018) | Historical student course grade data | Drop out of a course | Undergradeate students | 100 | SVM, FFNN, PESFAM, LOGIT_Act | SVM (62.50) | LOGIT_Act(97.13%) |