Hybrid AI-Driven Intrusion Detection: Framework Leveraging Novel Feature Selection for Enhanced Network Security

In today's rapidly evolving digital landscape, safeguarding network infrastructures against cyberattacks has become a critical priority. This research presents an innovative AI-driven real-time intrusion detection framework designed to enhance network security, particularly in Wireless Sensor Networks (WSNs) and Cloud Computing (CC) environments. The system employs classical machine learning models, Logistic Regression, Decision Tree, and K-Nearest Neighbors, optimized through the novel Energy Valley Optimization (EVO) method using the NSL-KDD dataset. Feature selection significantly reduced the number of input features from 42 to 18 while maintaining strong detection capabilities. The proposed system achieved 98.95 percent accuracy with Decision Tree, 98.47 percent with K-Nearest Neighbors, and 88.84 percent with Logistic Regression. Moreover, high precision, recall, and F1-scores were attained across all classifiers while substantially reducing training and testing times, making the framework highly suitable for real-time applications. To ensure fair detection across diverse attack types, dataset balancing via downsampling was applied to address class imbalance challenges. This investigation focuses on the significance of advancing intrusion detection systems in cloud computing and WSNs. Overall, this work advances secure communications by delivering a scalable, low-latency, and high-accuracy intrusion detection solution aligned with the latest trends in artificial intelligence, cybersecurity, and real-time digital networks