scientific article; zbMATH DE number 1264366

From MaRDI portal
Publication:4233295





The authors investigate the convergence property of the conjugate gradient algorithms which depend on the search directions \[ d_1=-g_1, \quad d_{k+1}=-g_{k+1}+\beta_kd_k,\quad k\geq 1, \] by using different choices for the scalar \(\beta_k\), where \(g_k\) is the gradient of \(f\) at \(x_k\). Under some assumptions which are slightly weaker than those proposed by \textit{Y. F. Hu} and \textit{C. Storey} [J. Optimization Theory Appl. 71, No. 2, 399-405 (1991; Zbl 0794.90063)], the authors prove a global convergence result which allows \(\beta_k\) to be set in a wider range than that which Hu and Storey aquired.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4233295)