一种基因表达数据的混合特征选择方法  

A Hybrid Features Selection Method for Gene Expression Data

作　　者：黄道斌[1] 叶明全[1] 张玲[2] 胡天寒[1] 杨利[1] 

机构地区：[1]皖南医学院计算机教研室,安徽芜湖241002 [2]皖南医学院生化教研室,安徽芜湖241002

出　　处：《宿州学院学报》2014年第8期73-76,110,共5页Journal of Suzhou University

基　　金：安徽省高校省级自然科学研究重点项目"面向肿瘤基因表达数据的特征选择与集成分类研究"(KJ2014A266)

摘　　要：针对肿瘤基因表达谱具有高维度、小样本的特点,提出一种面向基因表达数据的混合特征选择方法。首先利用Relief算法从原始基因集中获取10%左右与分类相关的基因作为相关基因子集,然后利用SVM-RFE算法对相关基因子集再进行选择,获取对分类影响较大的排序基因子集,最后利用Wrapper方法对排序基因子集进行选择,获得最佳基因子集。以支持向量机为分类器,对急性白血病、结肠癌肿瘤表达谱数据进行测试,实验结果表明所提方法时间开销较小,且取得较高的分类精度。According to the characteristics of the tumor gene expression profiles which includes high dimensional- ity and small samples, this paper proposes a hybrid features selection method for gene expression data. Firstly, relief algorithm to get about 10% of the original gene set which is related to classification as the relevant gene subset,and then is used SVM-RFE algorithm to obtain the sorted gene subset that has greater impact on the classification from the relevant gene subset. Finally,is used the Wrapper method to select the optimal gene sub- set from the sorted gene subset. Using support vector machine as classifier,is used acute leukemia and colon can- cer mieroarray datasets for testing, experimental results show that the proposed method costs less time, and achieves higher classification accuracy.

关 键 词：基因表达谱 特征选择 肿瘤分类 支持向量机 

分 类 号：Q78[生物学—分子生物学]