Title:DoS Attack Detection System Using Machine Learning


Authors:Radhika Ranawat, Yuvraj Pawar, Sonal Gupta, Niharika Soni, Vnadana Kate, Vinayak Sharma


Published in: Volume 3 Issue 1 Jan June 2026, Page No.374-382


Keywords:DoS Attack, Machine Learning, Cy ber Security, Denial of Service.


Abstract:New types of cyberattacks called Denial of-Service (DoS) assaults now pose an extremely significant risk to contemporary computer networks, often leading to major disruptions in service availabil ity and resulting in considerable economic damage through loss of productivity and revenue. Outdated system designs struggle to recognize new threats due to lack of adaptability; they frequently produce many unnecessary alerts. The study introduces an advanced system for detecting Distributed Denial of Service attacks through machine learning algorithms aimed at improving immediate cybersecurity measures in networks. A new mechanism intercepts real-time or archived internet data through an analyzer tool, analyzing it according to metrics like transmission frequency of packets, randomness in Internet Pro tocol headers, diversity among protocols used, and movements at specific ports. An algorithmic system utilizing supervision, namely a RandomForest clas sifier, undergoes training using annotated data sets for proficient identification of benign versus harmful network activity. A well-trained algorithm constantly watches for traffic on networks and sends notifications when it spots unusual behavior suggesting intrusion attempts. The experimental assessment reveals re markable precision in detecting targets while signifi cantly lowering error rates associated with incorrect identifications. This framework offers an adaptable, resourceful method for maintaining network security, guaranteeing continuous operation and reducing dam age caused by changing digital risks.


Download PDF