一种基于改进的遗传算法的癌症特征基因选择方法  

Cancer Feature Gene Selection Method Based on the Improved Genetic Algorithm

作　　者：范方云[1] 孙俊[1] 王梦梅[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《江南大学学报（自然科学版）》2015年第4期413-418,共6页Joural of Jiangnan University (Natural Science Edition)　

基　　金：国家自然科学基金项目(61170119)

摘　　要：提出了一种基于改进的遗传算法的癌症特征基因选择方法,对基本遗传算法的交叉和变异操作加以改进,利用改进后的遗传算法对基因微阵列数据进行特征选择。使用支持向量机对基因子集进行留一交叉验证。实验结果表明,改进后的遗传算法比基本遗传算法在选择特征基因时具有明显的优势。A cancer feature gene selection method based on the improved genetic algorithm is proposed in this paper.The feature genes are selected from gene microarray data using the improved genetic algorithm,which is obtained by improving the crossover and the mutation operation of the basic genetic algorithm. Then the leave-one-out cross validation( LOOCV) is conducted by using the support vector machine( SVM) for each gene subset. The experimental results show that,the improved genetic algorithm has obvious advantages compared with the basic genetic algorithm in feature gene selection.

关 键 词：遗传算法 特征基因 微阵列数据 支持向量机 留一交叉验证 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]