Archimedes Fire Hawk Optimization enabled feature selection with deep maxout for network intrusion detection

Abstract

Network intrusion detection systems (NIDSs) are fundamental for protecting computer networks. In this paper, the Archimedes Fire Hawk Optimization (AFHO) enabled Deep Learning (DL) is developed. At first, the Wireless Network is initialized, and then the recorded log file is passed through the pre-processing phase utilizing quantile normalization, and missing data imputation process. Next to this, the feature selection is done with the help of AFHO, where AFHO is the combination of Archimedes optimization algorithm (AOA), and Fire Hawk optimization (FHO). At last, the intrusion detection (ID) is accomplished by Deep Maxout Network (DMN), which is tuned employing the proposed AFHO. The performance measures such as precision, recall and F-measure based on training data, where the K- fold acquired 93.7 %, 97.7 %, and 95.6 %.