PCA for antisymmetric matrix

250 Views Asked by At

Principal Component Analysis (PCA) is often used to convert a symmetric matrix to lower-dimensional one. Is there a similiar method for antisymmetric matrices?

As we known, the eigenvalues of antisymmetric is in pairs with $+$, $-$ respectively. Could I choose the $\max | \cdot |$ pairs to do PCA?

Is there good referrence recommented?

Thank you.