An Advanced design of Intrusion Detection System using Machine learning Architecture

Abstract

The Internet is advancing over the years, computer networks and its applications are having an exponential growth. An increased chances of getting attacked and major risk of potential damage caused by it, therefore many Intrusion Detection Systems (IDS) and Intrusion Prevention Systems (IPS) are instantiated to analyze networks and prevent any threats. Due to the limited number of datasets, such intrusion detection systems cannot be deployed accurately as well as are providing full proof solutions. In this paper, a comprehensive evaluation regarding the network features and outline the issue pertained by CICIDS Datasets is presented. This paper also presents an advanced design of Intrusion Detection System which can be used to classify any given network flow and declare Malicious or Benign system. The proposed system can evaluate different real-world models using deep learning techniques resulting in better responses