Enhancing Cybersecurity Against DDOS Attacks Evaluating Supervised Machine Learning Techniques

Abstract

Listen

Translate article

An individual or group launches a cyber attack when they intentionally try to get into another person's or group's computer system. Typically, the goal of an attacker is to gain an advantage by interfering with the victim's network. Now that COVID-19 has wreaked havoc on businesses throughout the world, it's cybercriminals' ideal storm. When it comes to cyber threats, Distributed Denial-Of-Service attacks (DDoS) are the most common and dangerous for corporate networks, apps, and services. Distributed denial of service attacks aim to flood a server, service, or network with malicious traffic in an effort to interrupt regular traffic. Financial losses, decreased productivity, damaged brands, worse credit and insurance ratings, damaged relationships with suppliers and customers, and IT budget overruns are all possible outcomes. Developing Network Intrusion Detection Systems (NIDSs) that can reliably foretell DDoS attacks is an urgent issue. This study used the CICDDoS2019 dataset to assess supervised Machine Learning (ML) methods. The machine learning algorithms that were assessed include AdaBoost, Naïve Bayes, MLP-ANN, Random Forest, and SVM. We use the assessment metrics: Area Under the Curve (AUC), Accuracy, F-measure, Precision, and Recall. This study demonstrates that of the algorithms tested, AdaBoost shows the highest promise in detecting DDoS attacks