An interpretable IoT intrusion detection method based on a hybrid deep learning model
Shuqin Zhang, Shaoyun Liu · 2025
With the rapid development of the Internet of Things (IoT), network security threats are becoming increasingly serious. Intrusion detection system (IDS) is an important technology for ensuring IoT security. To ensure the security of the IoT, this study proposes a novel intrusion detection model based on a hybrid deep learning model. This model effectively improves the accuracy of intrusion detection by extracting spatio-temporal features of data in parallel. We conduct multi-classification experiments on the CICIDS2017 dataset, and the experimental results show that the accuracy of this method exceeds 99%, which is better than the traditional methods. This study also uses the SHAP method to explain the prediction results of the model, enhancing the transparency and credibility of the model's decision-making. This study proposes an efficient, accurate and interpretable intrusion detection method for IoT environments, which is beneficial for maintaining IoT security.