A comparative study of multi-objective optimization algorithms for sparse signal reconstruction

The development of the efficient sparse signal recovery algorithm is one of the important problems of the compressive sensing theory. There exist many types of sparse signal recovery methods in compressive sensing theory. These algorithms are classified into several categories like convex optimizati...

Full description

Saved in:
Bibliographic Details
Published in:The Artificial intelligence review 2022-04, Vol.55 (4), p.3153-3181
Main Authors: Erkoc, Murat Emre, Karaboga, Nurhan
Format: Article
Language:English
Subjects:
Algorithms
Artificial Intelligence
Comparative studies
Computational geometry
Computer Science
Convex analysis
Convexity
Knee
Mathematical optimization
Multiple objective analysis
Optimization
Optimization techniques
Recovery
Search methods
Signal reconstruction
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The development of the efficient sparse signal recovery algorithm is one of the important problems of the compressive sensing theory. There exist many types of sparse signal recovery methods in compressive sensing theory. These algorithms are classified into several categories like convex optimization, non-convex optimization, and greedy methods. Lately, intelligent optimization techniques like multi-objective approaches have been used in compressed sensing. Firstly, in this paper, the basic principles of the compressive sensing theory are summarized. And then, brief information about multi-objective algorithms, local search methods, and knee point selection methods are given. Afterward, multi-objective sparse recovery methods in the literature are reviewed and investigated in accordance with their multi-objective optimization algorithm, the local search method, and the knee point selection method. Also in this study, examples of multi-objective sparse reconstruction methods are designed according to the existing studies. Finally, the designed algorithms are tested and compared by using various types of sparse reconstruction test problems.
ISSN:0269-2821
1573-7462
DOI:10.1007/s10462-021-10073-5