Deep Learning Models in Computer Data Mining for Intrusion Detection

Abstract

In recent years, the expanded usage of wireless networks for the transfer of enormous amounts of data has caused a multitude of security dangers and privacy issues; accordingly, a variety of preventative and defensive measures, such as intrusion detection systems, have been developed. Intrusion detection techniques serve a crucial role in safeguarding computer and network systems; yet, performance remains a serious concern for many IDS. The effectiveness of IDS was analyzed by constructing an IDS dataset comprised of network traffic characteristics to identify attack patterns. Intrusion detection is a classification challenge requiring the use of Deep Learning (DL) and Data Mining (DM) methods to categorize network data into regular and attack traffic. In addition, the kinds of network assaults have evolved, necessitating an upgrade of the databases used to evaluate IDS. In this study, we present a deep learning-based IDS that combines an optimization technique called spider monkey swarm with a convolutional neural network. With the use of the well-known NSL-KDD dataset, the SMSO-CNN is assessed and contrasted with the following techniques: deep neural networks, k-nearest neighbor, and long-short team memory. The experimental findings demonstrate that the SMSO-CNN outperforms other approaches in terms of accuracy.