Evaluating the Efficacy of Machine Learning in Detecting DoS & DDoS Attacks: A Comprehensive Dataset Analysis

Abstract

Machine learning is currently being widely employed to create malware detection systems (MDS). These systems can detect and categorize various types of cybercrime, including DoS and DDoS attacks. These attacks may involve multiple unique IP addresses and machines, and can be facilitated by malware. DoS and DDoS attack scans are carried out on a daily basis, and non-profit multinational companies, regardless of size, can fall victim to them. Such attacks can cause a significant slowdown or even bring down the online services, email, websites, and other digital operations of these companies. Cybersecurity operations may sometimes be disrupted by DoS and DDoS attacks, which can allow criminal activities such as data theft and network infiltration to take place, resulting in the loss of valuable company data. Our DoS and DDoS dataset for the malware detection system employs a single methodology, which involves the use of Python code. This paper will focus on assessing the accuracy of the DoS and DDoS dataset within the malware detection system.