Clustering Analysis of Microarray Gene Expression Data with New Clustering Ensemble Method

摘要

DNA microarray experiment is an attracting technology which can monitor expression of thousands of genes at the same time. Using the promising technology, accurate classification of tumor subtypes at molecular level becomes possible. Meanwhile, it can help discover interesting gene expression patterns to extract underlying biological knowledge. Many machine learning methods such as clustering, classification and so on have been successfully applied. Especially, clustering as a basic unsupervised learning approach is widely used. However, there are still some shortcomings in clustering methods, e.g. automatically deciding the natural clustering number and improving clustering accuracy. Facing these shortcomings in clustering, in this paper, we propose a new clustering ensemble method to solve them, which is prior to what is achieved by a single clustering algorithm, and apply it to DNA microarray data analysis. The final experiment result shows that the new method works well.