A PCA decomposition of the original data can be written as
![]() |
(2) |
Note that this is the same as the ICA decomposition, but uses a different function to minimise -- in this case, one that measures correlation.
The PCA decomposition is easily found using SVD. That is
, with
and
being orthogonal matrices (i.e.
). Hence the PCA decomposition is given by
and
.