Artificial Immune System And Its Data Clustering Applications

Biological immune system is a highly parallel adaptive information learningsystem, which can identify and remove the antigenic eyewinkers invading the body.This system can learn, remember and adjust adaptively to keep the stabilization insidethe body. During recent years, people begin to realize the revelatory significance ofthe biological immune mechanism to intelligent algorithm. Artificial Immune System(AIS) is this kind of new algorithm which is inspired by the biological immunesystem. This kind of algorithm has been used in many fields, such as machinery study,unconventionality and malfunction diagnosis, simulation of the behavior of robots,control of robots, inbreak detection of networks and etc. It has been a new effectivemember of the family of intelligent algorithms.However, artificial immune system (AIS) is a new research area, the results arein disorder and short of systematization. Especially, some so-called AIS essentially aremodified forms of traditional methods, not really systems based immune system.Furthermore, both modeling and application of AIS are not so thorough and perfectthat the important characteristics of natural immune system are not fully realized andutilized.Now in the AIS application researches, the research about AIS applying to dataclustering is more and more. But some questions have produced during AIS applyingto data clustering. For example, some parameters must be input by ourself. Parametersare sensitive for network changing. Parameters in network are not enough to analyze.In this paper, some basic concepts, framework, functions and principles of thebiological immune system are introduced. Then the research works analyze somemain methods. Based on former works, put forward a common frame of AIS. Lastexited methods are analyzed and based on them, put forward a new algorithm, thealgorithm is validated by simulation study. In a word, the mainly result of our worksare:1. Put forward a common frame of AIS. 2. Then analyze the existing artificial immune algorithm for data clusteringaccording to the common frame of AIS. Propose a novel adaptive artificial immunealgorithm for data clustering. Analyze and optimize the parameters carefully anddeeply. Therefore decrease of dependence of the algorithm on the problems andalleviate the customer's burden because some important parameters can be adaptivelyobtained in the course of the operation. Last the algorithm is validated by simulationstudy.