176

R. O. Ogundokun et al.

efficiently delivers a high accuracy and detection rate (DT) and a low false-positive

rate (FPR). The formula of the three metrics are as follows [22]:

Accuracy:

T P + T N

T P + F P + F N + T N

(1)

DT:

T P

T P + F N

(2)

FPR:

F P

F P + T N

(3)

3

Findings and Discussion

The results discovered and the interpretation of the results are discussed in this

section.

3.1

Findings

Figure 2 displays the confusion matrix gotten during the execution of the T-KNN

classifier while Fig. 3 displays the confusion matrix for the I-KNN classifier. It was

discovered from the True Negative (TN), True Positive (TP), False Negative (FN),

and False Positive (FP) values that the I-KNN outperformed the T-KNN classifier.

4

Discussion

The system’s performance was assessed using a confusion matrix, which included

parameters such as accuracy, detection rate, and false-positive rate. Accuracy and

performance are both improved when the false-positive rate is reduced. In this work,

the limitations of various approaches of accuracy, such as false-positive rates, are

discussed [23]. Table 1 displays the confusion matrix for the classifiers, and Table

2 displays a comparison of the two classifiers used in this work, the T-KNN and

the I-KNN. The I-KNN surpassed the T-KNN in terms of accuracy and FPR with an

accuracy of 96.7% over 81.7% and an FPR of 0.048 over 0.22. The T-KNN surpassed

the I-KNN in terms of DR with 100% over 97.4%, respectively.