Red fox optimizer based feature selection with optimal deep learning based Intrusion detection for network security

Red fox optimizer based feature selection with optimal deep learning based Intrusion detection for network security

연구 분야: Networking

학회: International Journal of Information Technology

In recent times, increasing severity and complication of security assaults on computer networks are simulated security researchers to incorporate various machine learning approaches to safeguard an organization’s reputation data and. Deep learning (DL) is one exhilarating method which been widely used by the intrusion detection system (IDS) for training to enhance their efficiency in securing the hosts and computer networks. IDS offers the solution for different security-based problems with distinct kinds of intruders or malicious attacks in networks. This study introduces a new red fox optimizer-based feature selection with optimal deep learning-based intrusion detection (RFOFS-ODLID) for network security. The presented RFOFS-ODLID technique accomplishes network security via the identification and classification of intrusions. Primarily, the presented RFOFS-ODLID technique performs a data scaling process using a min-max normalization approach. To elect an optimal set of features, the RFOFS technique is used. For the identification and classification of intrusions, the RFOFS-ODLID technique exploits the gated recurrent unit-recurrent neural network technique with Adamax as a hyperparameter optimizer. The performance validation of the RFOFS-ODLID approach is inspected on a benchmark dataset and the outcomes were investigated concerning different measures. The experimental outcomes stated that the RFOFS-ODLID technique reaches superior performance over other existing techniques.