Automated P2P Learning in Agent-Based Classification Networks

Peer-to-Peer (P2P) computing is a novel computing paradigm receiving ever increasing attention of the research community. It provides new opportunities in design and implementation of large scale intelligent systems satisfying modern requirements to scalability, autonomy, mobility, fault tolerance, etc. This paradigm becomes particularly attractive if it is integrated with multi-agent systems. The paper is focused on P2P cooperative decision making and P2P machine learning of cooperation of autonomous agents in open P2P networks. It proposes P2P mechanism for decision combining utilized by autonomous decision making agents and for P2P learning of decision combining. The paper results are validated using case study, P2P agent-based intrusion detection system.