HID-AG: Elevating cybersecurity with HybridIDNet and PCA-anomalyguard for intrusion detection excellence

Anomalous patterns in network traffic remain a critical obstacle in modern cybersecurity. Conventional intrusion detection systems (IDS) frequently face limitations in recognizing subtle or adaptive anomalies that signal advanced cyberattacks. To address these challenges, this study presents HID-AG — an IDS specifically developed to enhance anomaly detection capabilities. At its core, HID-AG incorporates HybridIDNet, which synergizes the strengths of CNNs, RNNs, and RF. Within this framework, the CNN module specializes in extracting spatial features, enabling HID-AG to identify anomalies embedded within complex network traffic structures. Employing this HybridIDNet architecture, the proposed approach achieves a notable accuracy of 97.84%, underscoring its effectiveness in confronting intricate cybersecurity threats. Implemented in Python, HID-AG promotes both accessibility and transparency, offering cybersecurity practitioners a practical and adaptable tool for real-world network anomaly detection. Beyond detailing the technical architecture, this paper also delivers a thorough evaluation of HID-AG’s performance in addressing anomaly detection tasks.