Enhancing intrusion detection system performance under imbalanced data conditions using a hybrid deep learning framework

Key Points

• To improve the performance of intrusion detection systems when facing imbalanced data.
• Developed a hybrid deep learning framework for intrusion detection.
• Analyzed the performance under varying conditions of data imbalance.
• Implemented various techniques to enhance detection rates.
• Achieved higher accuracy in detecting intrusions compared to traditional methods.
• Demonstrated improved detection rates for minority classes in imbalanced datasets.

Abstract