基于支持向量机的中国地鼠分类特征基因选取

针对中国地鼠基因表达谱数据维数高和样本小的特点，提出一种基于支持向量机(SVM)的分类特征基因选取方法。该方法利用改进的Fisher判别(FDR)基因特征计分准则剔除分类无关基因，提出由空间距离和功能距离组成的新距离作为相似性度量的标准进行冗余基因的剔除，采用SVM作为分类器检验特征基因的分类性能。实验结果表明，该方法有效地剔除了分类无关基因和冗余基因，选取的特征基因满足对中国地鼠正确分类的最小基因数。

Feature gene selection for Chinese hamster classification based on support vector machine

Concerning the gene expression profile of Chinese hamster feature, such as high dimension and small sample, a method of feature selection for Chinese hamster classification based on Support Vector Machine (SVM) was proposed in this paper. The method used improved FDR gene feature score criterion to remove the genes irrelevant to the classification. A new distance composed by space distance and function distance was proposed as the criterion of comparability to remove redundant genes. A SVM was used as classifier to validate the classification performance of the feature genes selected. The experimental results show that this method effectively removes the irrelevant and redundant genes, and selected the feature genes that meet the needs of least feature genes which classify accurately on Chinese hamster.