Agent-based risk learning for computing systems
Xiao Feng Wang, Han Kiliççöte, P.K. Khosla, Chengqi Zhang · 2002
Excessive or inadequate security controls can impair the utility of an open computing system. The agent based open computing environment we propose balances the amount of security control to maximize the utility of the system. In the system, intelligent agents learn the risk of the users and then dynamically adjust the security policy to achieve better system utility.