Where:
· TN: True Negative,
· TP: True Positive,
· FP: False Positive
· FN: false Negative.
As shown in Figure 15. the Naive Bayes classifier is more effective when compared to other Basic classifiers in predicting number of injuries. As illustrated in Table 1. The error rates and accuracy of each classifier are shown, showing the accuracy for the Naïve Bayes is about 81.59%. For the Naïve Bayes it took 0.03 seconds to run compared to the other classifiers with KNN having the same time and Bayes Network being 0.22 seconds, and the most time was 6.84 seconds for the J48 Decision Tree.