An Enhanced Hybrid Intrusion Detection Based on Crow Search Analysis Optimizations and Artificial Neural Network

The continuous advancement of computer networks has given rise to grave concerns regarding security and susceptibility. Network administrators utilize intrusion detection systems (IDS) to deliver essential network security. Current IDS devices produce false alarms in response to routine user activities rather than detecting novel assaults. Neural networks may be used to overcome this problem and increase detection accuracy. In this paper, we propose a hybrid approach based on neural networks and correlation-based feature selection to detect anomalies. Experimental research is done on the standard dataset NSL-KDD for intrusion detection using current attacks. We introduce a novel hybrid crowd search analysis optimization with an artificial neural network (HCSAOANN). The findings demonstrate that it outperforms in high accuracy, precision, F1-Score, and recall. In the proposed HCSAOANN algorithm, to explore the feature space, we merged the crow search optimization (CSO), which can converge into the overall best solution in the search function, with the upgraded version of the crow search analysis method. Some performance criteria were applied in the studies using an artificial neural network (ANN) as a classifier. The HCSAOANN methodology outperformed as compared to the previous fuzzy technique and achieved 98% accuracy and a 98% precision rate, which is 2.2% better than the previous CSO-ANFIS technique and 8.87% superior to the FC-ANN technique.