|
| Research study | Main results |
|
| [28] | (i) SMO, NaiveBayesSimple, and BayesNet obtained the highest accuracy and F-measure. |
|
| [17] | (i) Decision tree models (REPTree) had a high prediction accuracy.
(ii) REPTree was less sensitive to missing values than J48. |
|
| [15] | (i) Neural network and Naive Bayes classification with SMOTE technique had a high accuracy with 75%.
(ii) Naive Bayes and neural network models produced almost similar accuracy level when the discretization method was applied. Decision tree had less accuracy for both methods. |
|
| [31] | (i) Three decision tree algorithms C4.5, CART, and ID3 were applied and the result indicated that C4.5 is the best classifier for prediction of student performance. |
|
| [29] | (i) Six classification algorithms considered were Naïve Bayes (NB), Unpruned Decision Tree5 (DT), logistic regression, support vector machine using an ANOVA kernal function (SVM), neural network (NN), and k-nearest neighbor (k-NN).
(ii) The results indicated that all algorithms had a good predictive accuracy for young students and KNN predicted very well for old students and the rest of the classifiers were poor. |
|
| [27] | (i) Two algorithms were used: J48 and random tree.
(ii) The result showed that random tree model was more accurate than J48. |
|
| [14] | (i) NBTree classification was performed with a pretty good accuracy. |
|
| [11] | (i) This study used C4.5, AODE, Naïve Bayesian, multi label k-nearest neighbor algorithms.
(ii) The result concluded that multilabelled k-nearest neighbor had the best accuracy among the others (C4.5, AODE, and Naïve Bayesian). |
| [20] | (i) Three algorithms were compared: C4.5, multilayer perceptron, and Naive Bayes.
(ii) Naive Bayes has a good prediction accuracy. |
|
| [4] | (i) Four classifiers were investigated: J48 DT, NB, SMO, and MLP.
(ii) The results show that J48 DT algorithm achieves the best performance compared to the other algorithms with an accuracy of 84.8%. |
|
| [2] | (i) BN, NB, SVM, C4.5, and CART are used to build the learning model to predict student performance.
(ii) SVM is the best classifier compared to the other BN, NB, C4.5, and CART. |
|
| [1] | (i) Two main algorithms, decision stump and J48, were applied.
(ii) J48 provides more accuracy. |
|
