The rapid proliferation of the Internet of Things (IoT) and the emergent of sophisticated cyber-attacks demonstrate the urgent need for complete, intelligent intrusion detection systems to guarantee security. Machine learning (ML) and deep learning (DL) models provide a strong groundwork for the development of such systems, as they have demonstrated an impressive performance in numerous applications, such as network intrusion detection. This chapter proposes an ML-DL approach based on the NF-UQ-NIDS-v2 dataset, which is novel and comprehensive. In this approach, ML algorithms were used for binary intrusion classification and DL algorithms were applied to perform multi-class classification, allowing for a finer-grained detection of different intrusion types. In the presence of class imbalance, the oversampling and data augmentation techniques were employed to ensure fair representation. The proposed hybrid ML-DL system demonstrates notable improvements in the accuracy and reliability of intrusion detection, representing a significant step forward for the field of cybersecurity.