A new solution for the least squares problem

DOI10.1080/00207169908804847MaRDI QIDQ4702135zbMATH OpenOpenAlexFDO

Publication date 10 August 2000

Published in International Journal of Computer Mathematics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1080/00207169908804847

generalized inverse Hilbert spaces iterative method least squares problem normal system

Equations and inequalities involving linear operators, with vector unknowns (47A50) Numerical solutions to equations with linear operators (65J10) Numerical solutions of ill-posed problems in abstract spaces; regularization (65J20)

Review text

For a given vector \(b\) and a continuous linear operator \(T\) find an element \(c\in E\) such that \(\|Tc- b\|= \inf_{x\in E}\|Tx-b\|\), where \(T\in{\mathcal L}(E,F)\) with Hilbert spaces \(E\) and \(F\). It is assumed that the range \(R(T)\) of the operator \(T\) is closed, \(R(T)= \text{cl }R(T)\). An iterative method to obtain the generalized inverse \(T^+\), such that \(c= T+b\) is a solution of the normal system \(T'Tx= T'b\) with operator \(T'\) adjoint to \(T\), is presented. Under the assumption that \(\dim R(T')= l\), \(l\) finite, it is proved that the generalized inverse \(T^+\) is attained with \(l\) iteration steps.

Recommendations

Cites work

Cited in

(7)

This page was built for publication: A new solution for the least squares problem

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4702135)