基于非参数方法的肿瘤基因表达数据挖掘

Data Mining for Tumor Gene Expression Profiles Using a Nonparametric Method

该文提出了一种基于非参数统计的模式识别方法,此方法并不对微阵列数据作总体分布假设,从而降低了噪声对预测结果的影响.该方法可适用于两总体及多总体的模式识别问题.通过对两个真实的肿瘤基因表达数据的分析,验证了方法的识别效果.

This paper presents a pattern recognition model based on a nonparametric method. The method does not rely on any specific distribution of the microarray data, so it can reduce the influence of noise on classification results. It is applicable to two-class and multi-class pattern recognition. The proposed method has been applied to two sets of real microarray data of various human tumor samples to verify the effectiveness.