Supervised AFRC (Ada boost fast regression) machine learning algorithm for enhancing performance of intrusion detection system
Abhishek Kumar Jain, Khushboo Tripathi · International Journal of Engineering & Technology · 2019
In recent wireless network play critical role in every activity of human life. This wireless network process sensitive data network communication requires appropriate cyber security. In order to offer cyber security in computer network antivirus, user authentication schemes, firewalls and access control techniques has been developed to detect abnormal activities and potential attacks in computer network. To ensure security Intrusion Detection System (IDS) is designed for network security. In this paper proposed a Adaboost Fast Regression Classifier for attack or malicious activity detection in IDS system. For analysis in this research used CICIDS 2017 dataset the main advantage of this dataset is redundant data are minimal hence accuracy of malicious detection is increased. Collected dataset is fed into MATLAB and evaluated with proposed AFRC mechanism. In proposed AFRC scheme AdaBoost classifier and regression classifier are combined for attack identification and classification. Comparative analysis of proposed AFRC scheme with existing approach exhibits significant performance in terms of attack identification with reduced computational cost.