Relief

  函数实现了lelief算法。（计算样本的每一个特征的分类权重，选出具有最大权重的一组特征。样本的特征就是基因）。 对权重较大的一组基因进行两两冗余分析。将相关性弱的和强相关中权重大的基因保留，即为特征基因。 已知基因表达谱数据、不同类样本的个数，自行设定分类特征的权重阈值。 对原始数据标准化的方法：（x－基因均值）/基因方差。 样本的类内类间距离采用欧氏距离。(Lelief function implements the algorithm. (Classification weight each feature samples is calculated weight of a selected set of features having the greatest weight. Characteristic is the gene sample). Of the weight of a larger set of pairwise genetic redundancy analysis. The significant correlation between weak and strong correlation in the right gene retention, that is characteristic gene. Number of known gene expression data, samples of different classes, wherein the classification set threshold own weight. Of raw data standardized methods: (x- Gene mean)/variance gene. The between-class distance class of the sample using the Euclidean distance.)