|
| Models | Predicted content | Predictive data sources | Prediction algorithms | Analysis of results | Prediction accuracy |
|
| Other models | Academic performance | Learning activity data, questionnaire data | Neural networks, structural equations | Neural networks outperform structural equations | Not explicitly stated |
| Learning performance | Gesture behavior data, login data | Machine learning, regression analysis | Learners’ learning preferences and cognitive abilities are better | 85.7% |
| Dropout rate | Learning behavior login data | Bayesian networks, neural networks, decision trees | Learning login behavioral data predicts dropout rates | 75%–85% |
| Learning success | Learning management system | Regression analysis | Learner prior knowledge predicts learner success performance | Unspecified |
| Learning persistence | Learning motivation, learning achievement | Logistic regression, decision trees, plain Bayesian | Plain Bayesian prediction outperforms logistic regression and decision tree prediction | 87.50 |
| Learning outcomes | Learning management system | Decision trees, plain bayesian | Learning outcomes correlate with focused engagement, frustration | Not specified |
|
| Based on the neural network model | Learning performance | Association rule mining | Neuronal data center optimization algorithm, RBF neural network | Neural network algorithms can make accurate and good predictions and can select predictive data by the strong correlation | 90.7% |
|
