Weighted low rank approximation and reduced rank linear regression
2004 (English)In: INT CONF ACOUST SPEECH SIG PR, 2004, 501-504 p.Conference paper (Refereed)
The weighted low-rank approximation (WLRA) problem is considered in this paper. The problem is that of approximating one matrix with another matrix of lower rank, such that the weighted norm of the difference is minimized. The problem is fundamental in a new method for reduced rank linear regression that is outlined here, as well as in areas such as two-dimensional filter design and data mining. The WLRA problem has no known closed form solution in the general case, but iterative methods have previously been suggested. Non-iterative methods that are asymptotically optimal for the linear regression and related problems are developed in this paper. Computer simulations, where the new methods are compared to one step of the well-known alternating projections algorithm, show significantly improved performance.
Place, publisher, year, edition, pages
2004. 501-504 p.
, International Conference on Acoustics Speech and Signal Processing (ICASSP), ISSN 1520-6149
Other Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-34986DOI: 10.1109/ICASSP.2004.1326304ISI: 000222174600126ScopusID: 2-s2.0-4644303658OAI: oai:DiVA.org:kth-34986DiVA: diva2:425318
IEEE International Conference on Acoustics, Speech, and Signal Processing Montreal, CANADA, MAY 17-21, 2004
QC 201106212011-06-212011-06-172012-03-01Bibliographically approved