DocumentCode :
455000
Title :
Polynomial Time and Stack Decoding Solutions to Bounded Error Subset Selection
Author :
Tewfik, Ahmed H. ; Alghoniemy, Masoud
Author_Institution :
Dept. of Electr. & Comput. Eng., Minnesota Univ.
Volume :
3
fYear :
2006
fDate :
14-19 May 2006
Abstract :
The goal of bounded error subset selection (BESS) is to find the sparsest representation of an Ntimes1 vector b using vectors from a dictionary A of size NtimesM, such that the approximation is within a distance delta from b. Here delta is a user defined approximation threshold. Specifically, the goal is to find the sparsest vector x such that parAx - bpar les delta. The BESS is a reformulation of the classical subset selection problem. We describe two enumeration approaches with bounded complexities that find the optimal solution to the BESS problem. In particular, the paper describes the first exhaustive enumeration solution to subset selection type problems with polynomial complexity. Furthermore, it also describes a lower complexity stack decoding approach that finds a solution to the BESS problem with a complexity that is proportional to that of orthogonal matching pursuit. The approaches described here have a markedly better rate-distortion behavior than any of the other known solutions to the subset selection and BESS problems
Keywords :
computational complexity; decoding; iterative methods; signal representation; bounded complexities; bounded error subset selection; exhaustive enumeration solution; orthogonal matching pursuit; polynomial complexity; polynomial time decoding solution; rate-distortion behavior; sparsest representation; stack decoding solution; user defined approximation threshold; Computer errors; Decoding; Dictionaries; Greedy algorithms; Iterative algorithms; Matching pursuit algorithms; Polynomials; Signal processing algorithms; Signal representations; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech and Signal Processing, 2006. ICASSP 2006 Proceedings. 2006 IEEE International Conference on
Conference_Location :
Toulouse
ISSN :
1520-6149
Print_ISBN :
1-4244-0469-X
Type :
conf
DOI :
10.1109/ICASSP.2006.1660592
Filename :
1660592
Link To Document :
بازگشت