Abstract
To cope with the growing number and changing nature of malicious cyber-attacks, machine learning techniques have been extensively employed to develop intrusion detection systems (IDS) for intelligent detection. However, these systems face numerous challenges, given the continually evolving nature of these attacks and their sheer volume. In here, a new hybrid intelligent IDS employing a combination of Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) is proposed. This model aims to classify audit data and predict potential security breaches by leveraging the strengths of both CNN and LSTM. The new hybrid model has shown a significant improvement in detection accuracy while reducing the detection time.