Aode-based Intrusion Detection with Subset Feature Selection using an Optimized Cascade Correlation Neural Network

This research introduces an Intrusion Detection System (IDS) that uses the Anonymised Online Data Expansion (AODE) classification approach to improve network threat detection. The system, written in Java and including a graphical user interface, analyses network traffic data in ARFF format for training and assessment purposes. The AODE-based model enhances anomaly detection by analysing network traffic patterns with a probabilistic learning method. It effectively identifies known and new assaults, supports both real-time and offline modes, and adapts to changing threats. The system's user-friendly interface enables the visualisation and understanding of findings, assisting security administrators in automated decision-making.Its scalable architecture allows it to fit into a large variety of network environments, ranging from business to cloud infrastructures. By integrating with current security frameworks, the IDS strengthens cybersecurity defences, seeking to decrease false positives while retaining high detection rates.