An Improved Correlation Measure-based SOM Clustering Algorithm for Gene Selection

摘要

Among the large amount of genes presented inmicroarray gene expression data, only a small fraction ofthem is effective for performing a certain diagnostic test.For this reason, reducing the dimensionality of geneexpression data is imperative. Self-organizing map (SOM) isa type of mathematical cluster analysis which particularlywell suited for recognizing and classifying features incomplex, multidimensional data. This paper proposes animproved Self-organizing map clustering algorithm whichbased on neighborhood mutual information correlationmeasure. To evaluate the performance of the proposedapproach, we apply it to six well-known gene expressiondatasets and compare our results with those obtained byother methods. Finally, the experimental results show thatthe proposed approach to gene selection is indeed efficient.