Accurate Deauthentication Attack Detection using SVM Classifier in Comparison with Decision Tree Classifier
DOI:
https://doi.org/10.47750/pnr.2022.13.S04.091Keywords:
Deauthentication attack, SVM, Novel Features Selection, Decision Tree Classifier, Machine Learning, EntropyAbstract
Aim: This study aims to compare the accuracy of Support Vector Machine Classifier to Decision Tree Classifier in detecting deauthentication attacks.
Materials and Methods: The dataset used in this study consists of 61,000 records. For testing, 9,604 records calculated using G power are used to achieve a 95 % confidence level in accuracy with 1% margin error. Each record consists of 42 attributes/features. In order to detect deauthentication attacks, SVM and Decision Tree are used.
Results: The accuracy of SVM was 87.02%, P<0.05, whereas the accuracy of the Decision Tree Classifier was 71.81%, P<0.05.
Conclusion: Comparing SVM to Decision Tree Classifier, the present study found that SVM performed significantly better in detecting deauthentication attacks.
