Development of an Intelligent Intrusion Detection Model using an Ensemble of Deep Learning Paradigm
Olamatanmi Josephine Mebawondu, Tajudeen Adeleke Badmos, Olufunso Dayo Alowolodu, Jacob Olorunshogo Mebawondu · 2024
The internet's advent has permeated every aspect of our lives, resulting in an explosion of information generation and management. Deep learning, as an Artificial Intelligence (AI) function, mimics human intelligence in data processing for decision-making. Deep learning (DL) encompasses a variety of machine learning techniques where unsupervised and supervised feature learning can be implemented across multiple layers in hierarchical architectures. Combining deep learning techniques can mitigate the weaknesses of individual methods. This research aims to integrate deep learning techniques to develop an intelligent intrusion detection model. The combined models yield a Network Intrusion Detection System (NIDS) with superior performance compared to individual techniques. The results demonstrate that the ensemble model achieves an accuracy of 79%, surpassing single deep learning models.