New Intrusion Detection Framework Using Cost Sensitive Classifier and Features
Phyo Thu Thu Khine, Htwe Pa Pa Win, Khin Nwe Ni Tun · International Journal of Wireless and Microwave Technologies · 2022
The huge increase amount of Cyber Attacks in computer networks emerge essential requirements of intrusion detection system, IDS to monitors the cybercriminals.The inefficient or unreliable IDS can decrease the performance of security services and today world applications and make the ongoing challenges on the Cyber Security and Data mining fields.This paper proposed a new detection system for the cyber-attacks with the ensemble classification of efficient cost sensitive decision trees, CSForest classifier and the least numbers of most relevant features are selected as the additional mechanism to reduce the cost.The standard dataset, NSL-KDD, IDS is used to appraise the results and compare the previous existing systems and state-of-the-art methods.The proposed system outperforms the other existing systems and can be public a new benchmark record for the NSL-KDD datasets of intrusion detection system.The proposed combination of choosing the appropriate classifier and the selection of perfect features mechanism can produce the cost-efficient IDS system for the security world.