Vol. 8, Issue 2 (2019)

Author(s):
Rajib Guha Thakurta

Abstract:
A machine learning method for identifying DDoS attacks is presented in this article. The limitations of traditional DDoS attack detection methods stem from their dependence on well-known attack signatures, which are easily circumvented by adversaries. On the other hand, machine learning algorithms have the ability to recognize patterns in regular network traffic and identify anomalies that might point to a DDoS attack. The suggested method achieves more accurate and dependable DDoS attack detection by combining random forests and decision trees. Tests conducted on actual datasets validate the effectiveness of the suggested approach, demonstrating that it outperforms conventional approaches in terms of accuracy and resilience to variations in attack patterns. For online companies and organizations that depend on the accessibility and security of their online services, this strategy has important ramifications. Researchers, practitioners, and students interested in DDoS detection and cybersecurity will find value in the study's findings.