There are two main techniques to implement PCA. The first technique, sometimes called classical, computes eigenvalues and eigenvectors from a covariance matrix derived from the source data. The second ...