Intelligent Recommendation System for Countering Network Attacks

This paper studies an approach to counteract network attacks based on network reconfiguration to eliminate the possibility of the successful completion of an attack by an intruder. To implement this approach, it is proposed to use a recommender system mechanism that provides both the generation of possible network topologies and their ranking. The proposed intelligent recommendation system is based on a reinforcement learning algorithm based on the actor-critic model. The conducted experimental studies confirm the effectiveness of the developed system.