An optimal and secure environment for intrusion detection using hybrid optimization based ResNet 101-C model

A monitoring system that can identify and assess abnormal activity is known as an intrusion detection system (IDS), and it is crucial in protecting the network against attacks. In an imbalanced dataset, the classification accuracy of the predictive model will decrease and the training time will be relatively long. The Hybrid Seagull Optimized ResNet 101-C (HSO-ResNet 101-C) technique is proposed in this study to precisely detect various attack types to overcome these issues. The computational complexity of the ResNet101-C architecture, which requires more time and resources when handling big data, is solved via the Hybrid Seagull Optimizer. To improve the performance of intrusion detection systems in big data environments some performance metrics such as accuracy, precision, recall, and F1-score are employed. In this paper, the two types of datasets CICIDS2017 and UNSW-NB15 can be utilized for detecting the intrusion detection rate. In the CICIDS2017 dataset, the performance rates of 100%, 99.88%, 98.9%, 99.2%, and 6.38% are obtained from the parameters of accuracy, precision, recall, and F1-score FAR respectively. The performance rate of accuracy, precision, recall, and F1-score FAR is 98.9%, 96.4%, 95%, and 97.8%, obtained from the UNSW-NB15 dataset.