Published Paper Details:

As cyber threats continue to grow in complexity and frequency, traditional intrusion detection systems (IDS) often struggle to keep pace, leading to high false positive rates and missed detections. This dissertation presents an AI-driven approach to optimizing IDS performance by leveraging machine learning and deep learning techniques to enhance detection accuracy, reduce false alarms, and improve adaptability to evolving attack patterns. The study explores multiple AI models, including Random Forest, Decision Trees, Logistic Regression, integrated with advanced optimization techniques such as feature selection and ensemble learning. Using benchmark datasets, the AI-optimized IDS achieved a significant performance improvement compared to traditional systems, with accuracy reaching 95.6%, precision at 94.2%, recall at 96.5%, and an F1 score of 95.3%. The results confirm that AI-enhanced IDS can offer more reliable and intelligent threat detection while minimizing false positives. This research contributes to the growing body of work on intelligent cybersecurity systems and lays the groundwork for future advancements in autonomous and adaptive intrusion detection.