HyDeck: Hybrid Decider-K for IoT Intrusion Detection System

The Internet of Things (IoT) is currently experiencing rapid growth in various sectors of human life, leading to significant advancements in technology and convenience. However, this rapid development is accompanied by security issues that pose threats to users. Traditional firewall technology falls short in identifying abnormal activities in network traffic, highlighting the necessity for an effective intrusion detection system (IDS). This research addresses this need by developing a model for intrusion detection that combines K-Means clustering, K-Nearest Neighbors (KNN), and Decision Tree algorithms to enhance performance on imbalanced datasets. To ensure the robustness of the model, the data is divided into several folds and subjected to a thorough cross-validation process. The results indicate that the developed model achieves better accuracy for each data fold compared to a standard decision tree classifier. This improvement underscores the significant potential of hybrid models in greatly enhancing the security and reliability of IoT networks.