| Current study |
1 cohort of 54 degrees in on-campus university |
5 ML classification methods |
See Tables 4–6
|
| [10] |
Several cohorts of 27 degrees in on-campus university |
Regression |
Explained variance: varying from 73.3% to 83.2%, depending on the area |
| [11] |
1 cohort of 54 degrees in on-campus university |
MANOVA |
Explained variance: 30% |
| [17] |
2 cohorts of 1 degree in on-campus university |
Regression |
Accuracy: 88.7% |
| [20] |
2 online courses |
Artificial Neural Networks |
83.6% and 87.3% for dropout and non-dropout classes, respectively |
| [22] |
5 cohorts of one on-campus university |
4 ML classification methods |
F not provided. Accuracy varying from 86.12% to 87.23%, depending on methods |
| [23] |
1 cohort of 3 degrees in on-campus university |
4 ML classification methods |
Not numerically but graphically provided. F: varying over 40% and below 80% methods |
| [24] |
1 online course |
5 ML classification methods |
F not provided. Accuracy varying from 78.17% to 83.89%, depending on methods |
| [25] |
4 online courses |
4 ML classification methods |
F not provided. Accuracy: 50%-94%. Recall: 20%-90%. Precision: 10.5%-81-8%. Depending on methods |
| [26] |
3 cohorts of online university |
3 ML classification methods |
F varying from 65.65% to 71.91%, depending on methods |
| [27] |
1 online course |
4 ML classification methods |
F not provided. Recall: 73.9% - 87%, depending on methods |
