Abstract: This paper introduces a model of clustering analysis of tumor gene data expression using wavelet for noise reduction. The model analyzes the gene expression data effectively using the methods of wavelet transfer, data mining and coupled two-way clustering. With this model, noise in the gene expression data can be reduced and accuracy of sample classification can be increased. The model is applied to the analysis of gene expression data of colon.

Key words: cluster analysis, gene expression data, specific expressed genes, wavelet transformation