Assessing DDoS Detection Accuracy through Semi-Supervised Techniques

Abstract

Despite the proliferation of advanced Machine Learning (ML) techniques in DDoS detection, this pervasive attack remains a significant menace to the Internet's integrity. Existing ML based DDoS detection methods fall into two categories: supervised and unsupervised approaches. This paper synthesizes insights from existing research endeavors, and enhance DDoS detection through machine learning methodologies, specifically focusing on semi-supervised techniques for analysis purposes. By harnessing the power of semi-supervised ML, we employ a succession of algorithms including Naive Bayes, Support Vector Machines (SVM), and Logistic Regression, focusing on factors crucial for detection accuracy. This paper mainly focuses on assessing various ML algorithms. In summary, this paper presents the potential of semi-supervised ML in augmenting detection accuracy. Keywords— DDoS detection, Machine Learning, SVM, Naïve Bayes, Logistic regression