Parallel Gene Clustering Using MapReduce

摘要

Data clustering has been considered as one of the most important techniques for unsupervised learning in diverse applications. Gene clustering is to find out groups of genes similarly expressed in large size of microarray data. Meanwhile, recent development of microar-ray technology generates a very large number of microarray data with low cost and handles more than 10,000 genes simultaneously in one chip. Thus, high performance computing of gene clustering has become increasingly important in microarray data analysis. In this paper, we propose a scalable parallel gene clustering method using the MapReudce programming model. The proposed method utilizes the k-means algorithm for identifying similar groups of genes. Experiment results show that the proposed method can offer good scalability with data size increases, and different numbers of nodes, and it can also provide effective clustering results against real microarray data.