Visualization and classification of microarray gene data by nonnegtive matrix factorization

Gene microarray technology is an effective tool to collect the expression levels of thousands of genes from a single array. However, exploitation of the huge amount of data generated by microarrays is difficult because they are complex and noisy high-dimensional data. In this work, we present a biclustering method nonnegtive matrix factorization (NMF) to reduce the dimensionality of the data and discover the underlying biological process from gene expression data of Alzheimer's disease (AD). The simulation results show that the reduction of dimension and identification of informatively biological process are useful for both visualization and analyzing of such high-throughput gene dataset.