基于SAM和GA/SVM的肿瘤基因表达谱分类算法

The Approach for Classification of Tumor Gene Expression Data Based on SAM and GA/SVM

基因芯片技术在肿瘤分型分类的研究中得到了广泛的应用.为了处理肿瘤基因表达谱数据,建立肿瘤分类预测模型,文中采用基因表达差异显著性分析方法,支持向量机,遗传算法相结合的多步骤降维分类方法.采用该方法处理大肠癌和白血病数据集,筛选到基因数量较少并且分类准确度较高的特征基因子集.实验结果表明,文中的方法可以快速有效地筛选肿瘤特征基因,获得更好的分类效果.

Microarray technology has been widely used in the research of tumor classification. To analyze gene expression data and establish prediction models of tumor classification, a multi-step gene dimensionality reduction method is proposed. The method combines with significant analysis of microarrays （SAM）, genetic algorithm （GA） and support vector machine （SVM）. The approach is evaluated over the two datasets of colorectal cancer and leukemia, a tumor related subset with small gene numbers and high classification accuracy is selected. The experiment results show the efficiency and effectiveness of the proposed method for screening tumor related genes and achieving better classification accuracy.