HYBRID INTRUSION DETECTION WITH DEEP FEATURE EXTRACTION AND ML CLASSIFICATION

Abstract

With the advancement of the digital network, there has been a constant increase in the number of instances of intrusions. In today’s world of cybersecurity, intrusion detection is alarmingly of great concern due to the serious consequences it can present. In the past few years, diverse “Machine Learning” (ML) and “Deep Learning” (DL) approaches have been employed for intrusion detection. However, regarding accuracy, both types of learning techniques have their limitations. A combined approach, however, presents a situation whereby much more research has to be conducted in order to determine its effectiveness in the detection of intrusions. In this work, a model was proposed that incorporates DL feature extraction with ML classification. It is, therefore, with high precision that this method recorded an astounding 96 percent accuracy in detecting intrusion. This paper has provided evidence that the model proposed in this paper outperforms the previous models, including DL and ML, in intrusion detection. These are likely to provide significant contributions to the improvement of cybersecurity.