Rapid progress in deciphering the biological mechanism of Alzheimer's disease (AD) has arisen from the application of molecular and cell biology to this complex disorder of the limbic and association cortices. The precise diagnosis of AD, however, has little progress and is also a challenging task. In this study, we investigate the DNA gene expression data of AD based on independent component analysis (ICA) to find significant genes for AD diagnosis. ICA exploits higher-order statistics to identify gene expression profiles as linear combinations of elementary expression patterns that may be interpreted as potential regulation pathways. This method can identify many genes and related pathways that play a prominent role in AD and relate the activation patterns of these to AD phenotypes. Using the extracted significant genes, the classification of AD and control samples gets more easy and effective by less gene data. This report shows that ICA as a microarray data analysis tool could help us to understand the phenotype-pathway relationship and, thus will help us to elucidate the molecular taxonomy of AD.