Fig 1.
Workflow diagram for analyzing the key factors that influence the school dropout rate.
Table 1.
Description of categorical independent variables or features.
Table 2.
Description of dependent variable or feature.
Fig 2.
Confusion matrix.
Table 3.
Summary of Encoding Techniques Applied to Dataset Variables.
Fig 3.
Violin plot for numerical feature age.
Fig 4.
Countplot for categorical features.
Table 4.
VIF Scores Supporting Low Multicollinearity.
Table 5.
Statistical Analysis of Features Influencing School Dropout.
Table 6.
XGB 10-Fold Cross-Validation.
Table 7.
Evaluation of ML models for student dropout prediction.
Fig 5.
AUC-ROC curve for different ML models.
Fig 6.
A bar chart of performance metrics comparing ML models to predict school dropout students in Bangladesh.
Fig 7.
SHAP waterfall plot showing feature influence on student’s school continuation prediction.
Fig 8.
SHAP waterfall plot showing feature influence on student’s school dropping out prediction.
Fig 9.
LIME Analysis of Variable Impact on the Prediction of Continuing Education.
Fig 10.
LIME Interpretation of Variables Influencing Dropout Prediction.
Fig 11.
Data-Driven Framework for School Dropout Prediction and Prevention Policies.
Table 8.
Comparison of student dropout prediction models’ performance and methodologies with previous works.