Principal Component Analysis finds sequences of linear combinations of the features. The first linear combination maximizes the variance of the features (subject to a unit constraint). Each of the following linear combinations maximizes the variance of the features in the subspace orthogonal to that...
