• Two pattern recognition techniques PCA and ICA are studied and applied to the data classification problem

• PCA works better than K-mean classification

• Since ICA doesn't required the transform vector to be orthogonal, ideally it will give better result than PCA does