Dynamic self-defined immunity model base on data mining for network intrusion detection

Artificial immunity model (AIM) is a good approach to realize intrusion detection. In AIM normal data set (i.e., don´t contain attacks codes) is necessary to define self, before the model can be used. However, it is difficult to automatically get clear data set in practice. In the paper, we propose a novel dynamic self-defined immunity model which combine data mining techniques to improve the exist model. The self in the new model can be automatically defined and updated to adapt normal changes of network.