Research On Negative Selection Algorithm Of Artificial Immune System

Biological Immune System is a system with the ability of pattern recognition, distributed detection, memory capacity, diversity, overall learning ability, etc. Many of these abilities are desired by the system that solves the practical application and project.Artificial Immune System is a new discipline based on the features of immune system to solve the problems encountered in practical application and project. Negative selection algorithm is a bionics algorithm that references the negative selection principle of Immune System, is one of the core algorithms Artificial Immune System developments. Currently, negative selection algorithm is widely used for pattern recognition, virus detection, network intrusion detection, anomaly detection and some other specific applications.Based on the study of Forrest negative selection algorithm, a new algorithm is approached to improve the existing algorithm which has larger mature detectors set and large number "hole" in the pattern space. Then, the effectiveness of this new algorithm and parameters decision is analyzed.The study of this paper summarized as follows:1. The study of the parameters' impact to Forrest negative selection algorithm is described in detail, and points out the factors that needed to be improved in the existing algorithm.2. Puts forward a threshold adjustable negative algorithm based on Hamming distance matching rule, and analyses the decision of parameters of this new algorithm.3. The new algorithm is analyzed and tested with simulation experiments. Compared with the existing algorithm, the results show that the new algorithm has a less mature detectors set, therefore is faster than the original algorithm, and that new algorithm effectively reduces the number of "hole" in the space.4. The research work in this paper is summarized, and discusses the next phase of work.