A hybrid approach for feature selection using SFS with extra-tree and classification using AdaBoost with extra-tree

Cyberattack is a new trend in data security. Intrusion detection systems gather different information from PCs and networks to recognise security risks and evaluate an attack's information. The discovery rate is low despite an enormously large amount of data. The problem can be overcome by feature selection. In this research study, the proposed model contains a hybrid model for intrusion detection. This research shows hybrid model involves three sections: preprocessing, feature selection, and classification. We combined sequential forward search (SFS) and an extra-tree algorithm for feature selection. For classification, we combined AdaBoost and extra-tree algorithm. The proposed model was implemented on various datasets like KDD'99, NSL-KDD, UNSW-NB15, CICIDS2017, and CICIDS2018 datasets. The proposed model evaluated parameters are false alarm rate, detection rate, and accuracy. The comparative result study displays that the proposed model defeats the existing algorithm.