earticle

영어

This paper proposes a fast immune recognition model based on immune response. Inspired by biological immune system, AIS for anomaly detection has been adopted widely because of its analogy with body resistance in the immune system provided against agents which causes diseases. This paper mainly studies correspondence between immune response and anomaly detection. Compared with traditional AIS, the change of antigen type is judged by statistical techniques, which ignores the differences of the different systems. Antibody recognition is initiated by the proposed model only when the type of antigen has changed, which improves the efficiency of the algorithm. Complexity analysis shows the proposed algorithm is a linear algorithm. The usefulness of the proposed model is demonstrated through experiments. The experiments illustrate the availability and feasibility of the model.