Effective detector set generation and evolution for artificial immune system

Human bodies defend themselves against harmful invaders by the Natural Immune System (NIS). The NIS possesses a huge set of detectors against many different invaders. Based on the essential principles of the NIS, the idea of Artificial Immune System (AIS) can be used for protecting computer resources. The AIS can detect the nonself-activity using a self-adopted detector set. The detector set is generated randomly and evolved by a Genetic Algorithm (GA). However, the random generation and the blind GA adoption decrease the performance of detector set and generate unnecessary redundancy. To overcome this problem, we propose a more efficient detector generation and evolution scheme. Using the proposed idea, the AIS can produce more effective detector set with time advantages.