基于演化超网络的DNA微阵列数据分类方法

为能够更好地从高特征维度的DNA微阵列数据中挖掘癌症相关基因,实现对恶性肿瘤的分子分型,提出了一种基于演化超网络模型的DNA微阵列数据分类方法?演化超网络是受生物网络启发而建立的一种认知学习模型,其学习过程非常适用于发掘基因间的相互作用?该方法采用信噪比进行基因选择,选择后的基因经归一化后用于演化超网络的学习和分类?通过急性白血病和结肠癌2种数据集进行实验,结果表明,演化超网络在分类精度方面与当前其他方法有较高的可比性?

DNA microarray data classification based on evolutionary hypernetworks

With the purpose of finding the cancer-related genes from the high-dimension DNA microarray data for the classification of tumors,an evolutionary hypernetwork model is proposed in this paper.The evolutionary hypernetwork is a kind of cognitive learning model inspired by the biological networks and its learning process is very suitable for mining gene-geneinteractions.In this paper,the gene selection is based on a signal-to-noise ratio method,and the selected normalized genes are processed by the hypernetwork through the learning and the classification phases.Empirical studies on the acute leukemia dataset and the colon cancer dataset demonstrate that the proposed hypernetwork classifier is high comparable with other state-of-the-art approaches in terms of classification rate of learned results.