New perspectives on optimal transforms of random vectors

Date

2009

Authors

Howlett, P.
Pearce, C.
Torokhti, A.

Editors

Pearce, C.
Hunt, E.

Advisors

Journal Title

Journal ISSN

Volume Title

Type:

Book chapter

Citation

Optimization: Structure and Applications, 2009 / Pearce, C., Hunt, E. (ed./s), vol.32, pp.245-259

Statement of Responsibility

P. G. Howlett, C. E. M. Pearce and A. P. Torokhti

Conference Name

Abstract

We present a new transform which is optimal over the class of transforms generated by second-degree polynomial operators. The transform is based on the solution of the best constrained approximation problem with the approximant formed by a polynomial operator. It is shown that the new transform has advantages over the Karhunen–Loève transform, arguably the most popular transform, which is optimal over the class of linear transforms of fixed rank. We provide a strict justification of the technique, demonstrate its advantages and describe useful extensions and applications

School/Discipline

Dissertation Note

Provenance

Description

© 2009 Springer. Part of Springer Science+Business Media

Access Status

Rights

Copyright 2009 Springer Science+Business Media

License

Grant ID

Call number

Persistent link to this record