Fig 1.
Feature distribution histogram, where 0 = low, 1 = medium and 2 = high.
Table 1.
Dataset description.
Fig 2.
Feature distance calculation using a dendrogram graph.
Fig 3.
Correlation matrix heatmap.
Table 2.
Dataset distribution between train and test sets.
Table 3.
Feature importance values extracted using RF.
Fig 4.
Decision boundary plotting for SVM.
Fig 5.
LIME explanation of a "High" risk class.
Fig 6.
Decision tree explanation.
Fig 7.
One tree is represented by the random forest model.
Fig 8.
A confusion matrix shows the classification result (in %) of test data using four ML models.
Table 4.
Detailed results of four ML models on the overall test dataset.
Fig 9.
Learning curves for four models before parameter tuning.
Table 5.
Comparison of results due to parameter tuning for five k-fold cross-validation.
Fig 10.
Learning curves after parameter tuning for four models.
Table 6.
Result comparison among literature and this work.