The Role of Artificial Intelligence in Enhancing Security for Computer Networks

Abstract

This research proposes a novel AI-based framework with the purpose of improving network security by combining anomaly detection by means of Feedforward Neural Networks (FNNs) and dynamic threat response by Reinforcement Learning (RL). The framework is based on a four-tier conceptual model of monitoring, feature extraction, AI analysis, and response action execution. The FNN is used for activity categorization and abnormality identification to accurately determine security threats; RL is used for real-time decision making with alert notifications and traffic blocking depending on the status of the network. The empirical analysis proves the effectiveness of the proposed framework, which obtained the accuracy of 96.5% and the cumulative average RL reward of +12.5, which indicates the ability to reduce false positive and focus on important actions. The features of the scalability and adaptability of the proposed framework were analyzed, which proved its effectiveness in addressing modern threats. This research contributes to the AI-based cybersecurity research by proposing a scalable and real-time solution to the existing gap between threat identification and dynamic response, which creates a robust defense system against new cyber threats.