Low-Rank Matrix Approximation and Subspace Tracking
The usual way to compute a low-rank approximant of a matrix H is to take its singular value decomposition (SVD) and truncate it by setting the small singular values equal to 0. However, the SVD is computationally expensive. Using the Hankel-norm model reduction techniques in chapter 10, we can devise a much simpler generalized Schurtype algorithm to compute similar low-rank approximants. Since rank approximation plays an important role in many linear algebra applications, we devote an independent chapter to this topic, even though this leads to some overlap with previous chapters.
KeywordsSingular Value Decomposition Elementary Rotation Subspace Estimate Subspace Tracking Principal Subspace
Unable to display preview. Download preview PDF.