Detection of Denial-of-Service (DoS) Attacks Using Machine Learning: Classification and Performance Evaluation

Abstract

Cloud computing paradigm is we can pay by use only, which will result to ensure more data elements are getting used by the machine. This results, not to worry the unnecessary services which are not in use. In our model, will discuss about the orchestrated methods in stealthy denial of service strategies. Main concern, about this is it slowly increases the provider intensity, where no one will be able to notice it. This will make the mechanism to switch between servers, DoS cyber-attack targets the legitimate users for obtaining the accesses of resources and attacks the services by flooding the target with a huge traffic. Multiple ways to provide protection, in our research we will discuss all the significant ways in identifying the upfront Denial of Service attacks. Machine learning methods which include the cloud computing results the promising outcome by 99% in defending and detecting the cyber-attacks.
Keywords: Cloud Computing, Stealthy false data injection, cyberattacks classification, machine learning, network security, Random Forest.
Problem Definition: The current aim of this research study to develop the machine learning based classification which can effectively detect the Denial-of-Service (DoS) attacks by analyzing the network traffic patterns. To achieve the accurate classification the dataset will undergoes the many preprocessing techniques, including feature scaling and one-hot encoding of categorical attributes. This model aims to enhance the detection using machine learning methods in identifying and mitigating the potential threats in real world network environments.