ANALYSIS OF DOS ATTACK USING MACHINE LEARNING

Abstract

The use of the internet has increased significantly in today's digital world; however, this has also increased the potential risk of a denial-of-service attack. A DoS attack occurs when a malicious user tries to consume an excessive amount of computing and network resources, avoiding reasonable users from accessing them. The attacks can be triggered from any location and level of OSI model e.g network layer, transport layer, and application layers. The goal of the paper to identify DoS attacks using algorithm of Machine Learning and Neural Network, while focusing on application layer attack detection except transport and network layer. The experiments perform different train test split dataset. The experiment used the most recent DoS attack dataset, which was divided into different splits. The best decision tree and logistics regression split, it was discovered that Decision tree outperformed Logistics regression in terms of algorithms