基于并行的非支配排序遗传Ⅱ算法优化双聚类

Optimization Biclustering Algorithm Based on Parallel Non-Dominated Sorting Genetic AlgorithmⅡ

双聚类是微阵列基因表达数据分析中很实用的一种数据挖掘技术,它是一种同时对微阵列基因和条件进行聚类的方法,用来挖掘基因子集在条件子集下所体现出来的生物模式。传统的双聚类算法对于庞大的基因表达数据处理效率很弱,考虑在j Metal平台上实现基因表达数据的双聚类的一种新的研究方法及思路。同时考虑加入并行策略,提高算法的效率。在酵母啤酒细胞基因表达集和人类B-细胞两个标准数据集上对两个算法进行实验验证,表明所提出算法比其他多目标双聚类算法呈现出更好的优越性。

Biclustering is a very practical data mining technique in microarray gene expression data analysis and it is a way to cluster both microarray genes and conditions simultaneously, which is used to excavate the biological mode reflected by the gene subset set under the condition subset. The processing efficiency of traditional bielustering algorithm for large gene expression data is low, so this paper explores a new research method and idea, i.e. applying gene expression data bielustering on jMetal platform. Also the parallel strategy is proposed to improve the efficiency of the algorithm. Experiments on two datasets, yeast cell dataset and human B-cell dataset, show that our approach exhibits better and more stable performance than other multi-objective bielustering algorithms.