Result: A relaxed projection method using a new linesearch for the split feasibility problem
Title:
A relaxed projection method using a new linesearch for the split feasibility problem
Source:
AIMS Mathematics, Vol 6, Iss 3, Pp 2690-2703 (2021)
Publisher Information:
American Institute of Mathematical Sciences (AIMS), 2021.
Publication Year:
2021
Subject Terms:
Optimization, Composite material, Inverse Problems in Mathematical Physics and Imaging, Economics, Inverse Problems, Matrix (chemical analysis), 0211 other engineering and technologies, cq-algorithm, Geometry, 02 engineering and technology, Fixed-Point Problems, split feasibility problem, Mathematical analysis, Quantum mechanics, 01 natural sciences, QA1-939, FOS: Mathematics, Nonexpansive Mappings, Iterative Algorithms, 0101 mathematics, Mathematical Physics, Economic growth, Eigenvalues and eigenvectors, Physics, hilbert space, Projection (relational algebra), Mathematical optimization, Hilbert space, Iterative Algorithms for Nonlinear Operators and Optimization, Applied mathematics, Computer science, Materials science, Algorithm, signal recovery, Computational Theory and Mathematics, Computer Science, Physical Sciences, Convergence (economics), Computation, Inverse problem, Inverse, linesearch, Mathematics
Document Type:
Academic journal
Article<br />Other literature type
ISSN:
2473-6988
DOI:
10.3934/math.2021163
DOI:
10.60692/by1h3-p6t22
DOI:
10.60692/v8hyc-9bg76
Access URL:
Rights:
CC BY
Accession Number:
edsair.doi.dedup.....f12df48dee93ac92d07eba95a044ca2f
Database:
OpenAIRE
Further Information
In this work, we propose a new relaxed projection algorithm for the split feasibility problem with a new linesearch. The proposed method does not require the computation on the matrix inverse and the largest eigenvalue of the matrix. We then prove some weak convergence theorems under suitable conditions in the framework of Hilbert spaces. Finally, we give some numerical examples in signal processing to validate the theoretical analysis results. The obtained results improve the corresponding results in the literature.