Further Studies of the Application of Constrained Minimization Methods to Fredholm Integral Equations of the First Kind

From MaRDI portal
Publication:3683498





\textit{E. Babolian} and \textit{L. M. Delves} [J. Inst. Math. Appl. 24, 157- 174 (1979; Zbl 0428.65065)] proposed to use the Galerkin method with Chebyshev polynomials as basis functions to treat ill-posed problems \[ \int^{b}_{a}f(y)k(x,y)dy=g(x),\quad a<x<b. \] By truncating the Fourier-Chebyshev series they obtained a system of linear equations for which they minimized the residual in the \(l_ 1\) norm under the condition that upper and lower bounds of the solution are prescribed. The present author investigates use of the \(l_ 2\) norm, and he displays theoretical and heuristic considerations how to choose \(C_ f\) and \(r>\) in the regularizing condition \(| a_ j| \leq C_ fj^{-r}\), the \(a_ j\) being the Fourier-Chebyshev coefficients. They compare with the \(l_ 1\) algorithm and present a careful documentation of several numerical case studies.











This page was built for publication: Further Studies of the Application of Constrained Minimization Methods to Fredholm Integral Equations of the First Kind

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3683498)