基于支持向量机的肿瘤基因数据挖掘  

Data Mining for Tumor Gene Expression Data Based on Support Vector Machine

作　　者：于彬[1] 

机构地区：[1]青岛科技大学数理学院,青岛266061

出　　处：《科学技术与工程》2009年第24期7370-7372,共3页Science Technology and Engineering

基　　金：国家863高技术研究发展计划(2006AA02Z190);山东省教育厅科研基金(J06P04)资助

摘　　要：针对两种类别的肿瘤分类问题,首先运用信噪比方法筛选出表达水平发生显著性变化的特征基因,然后采用支持向量机作为分类器进行肿瘤分类,通过对两种类别的白血病DNA微阵列数据进行计算,达到了97.1%的分类准确度。To solve the problem of two-class tumor classification, this process is realized by selecting significant differentially expressed genes by signal to noise ratio and taking the support vector machine as the classifier to classify tumor. By using the method, the classification accuracy of 97.1% was obtained to Leukemia DNA microarray.

关 键 词：DNA微阵列 支持向量机 分类 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]