Singular Value Decomposition (SVD) #\[ A = U\Sigma V^T \]SVD works for any matrix and is central to PCA and low-rank approximation.Home | Matrix Decompositions