Hybrid Detector based Negative Selection Algorithm

摘要

A new algorithm of detector generation for negative selection algorithm is introduced by adding a big detector to reach a high coverage of non-self space. While the big detector can be variable in different shape using this concept, the paper puts forward an algorithm using a ring-hyper-sphere shaped detector as a big detector. The algorithm is tested using different self set (or training set) and real-word dataset Preliminary results demonstrate that the new approach enhances the negative selection algorithm in efficiency and reliability without significant increase in complexity.