Principal component analysis
= Principal component analysis
{wiki}
Principal component analysis finds orthogonal directions of greatest sample variance. Its first loading vector is a unit eigenvector corresponding to the largest eigenvalue of the sample covariance matrix.