The Research of Generation Algorithm of Detectors in Immune-Based Detection Model

In this paper, as the generation of naive B-cell in natural immune system, the concept of minimum set of candidate detectors (M) is introduced to cover the whole area of ID problems completely. Then the mechanism of initializing and storing set M are analyzed and provided. Based on M, a new negative selection algorithm (M-based Negative Selection Algorithm, MBNSA) is proposed. As compared with traditional algorithms, MBNSA generates detectors with higher efficiency. Additionally, to the problems arisen by blind and random detectorpsilas generation, two improved methods of MBNSA is put forward to realize the mechanism "detecting only when need". Experimental results demonstrate the efficiency of the proposed schemes.