Comparative study of PCA and ICA in the field of Data Reduction.

This paper presents the results of a comparative study of Pca and Ica in the field of data reduction. In particular, we compare the two feature extraction techniques- independent component analysis (ICA) and Principal component analysis (PCA) to project microarray data into statistically independent components and genes are clustered according to their mean distances from the calculated centroid. We test the statistical significance of enrichment of gene annotations within clusters. Result shows PCA outperforms ICA in constructing functionally coherent clusters on microarray Breast Cancer Wisconsin, Primary Tumours,, Parkinson’s tele monitoring and ecoli data and hepatitis data set