一种免疫否定分段匹配选择的数据分类方法

根据免疫否定选择原理，设计了基于掩码分段匹配的否定选择分类器，克服连续r位匹配法的缺陷。给出了适用于免疫优化的分类规则编码及分类信息分的评价。通过免疫进化对其进行群体优化以约简数据规则集。避免了传统分类算法缺乏全局优化能力的缺点，提高了对样本的识别能力。实验结果表明本文方法提高了数据分类的准确性，在数据分类准确率及平均信息分上优于传统的分类方法。

A New Method for Data Classification Based on Immune Negative Selection of Section Matching

Based on immune negative selection principle,a novel mask piecewise matching negative selection classifier is introduced to accomplish the date classification.In order to improve the accuracy of classified mining,a classification rule coding is defined and a criterion of information grade was proposed to obtain the rules.This method has a better command of obtain the global optimum than traditional algorithm.The simulation results illustrate that the immune classifier is an available and feasible algorithm for data classification,improved the classification’s accuracy and validity.