Volume 23, Issue 4
Convergence Properties of Multi-Directional Parallel Algorithms for Unconstrained Minimization

Cheng-Xian Xu & Yue-Ting Yang

DOI:

J. Comp. Math., 23 (2005), pp. 357-372

Published online: 2005-08

Preview Full PDF 22 886
Export citation
  • Abstract

Convergence properties of a class of multi-directional parallel quasi-Newton algorithms for the solution of unconstrained minimization problems are studied in this paper. At each iteration these algorithms generate several different quasi-Newton directions, and then apply line searches to determine step lengths along each direction, simultaneously. The next iterate is obtained among these trail points by choosing the lowest point in the sense of function reductions. Different quasi-Newton updating formulas from the Broyden family are used to generate a main sequence of Hessian matrix approximations. Based on the BFGS and the modified BFGS updating formulas, the global and superlinear convergence results are proved. It is observed that all the quasi-Newton directions asymptotically approach the Newton direction in both direction and length when the iterate sequence converges to a local minimum of the objective function, and hence the result of superlinear convergence follows.

  • Keywords

Unconstrained minimization Multi-directional parallel quasi-Newton method Global convergece Superlinear convergence

  • AMS Subject Headings

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address
  • BibTex
  • RIS
  • TXT
@Article{JCM-23-357, author = {}, title = {Convergence Properties of Multi-Directional Parallel Algorithms for Unconstrained Minimization}, journal = {Journal of Computational Mathematics}, year = {2005}, volume = {23}, number = {4}, pages = {357--372}, abstract = { Convergence properties of a class of multi-directional parallel quasi-Newton algorithms for the solution of unconstrained minimization problems are studied in this paper. At each iteration these algorithms generate several different quasi-Newton directions, and then apply line searches to determine step lengths along each direction, simultaneously. The next iterate is obtained among these trail points by choosing the lowest point in the sense of function reductions. Different quasi-Newton updating formulas from the Broyden family are used to generate a main sequence of Hessian matrix approximations. Based on the BFGS and the modified BFGS updating formulas, the global and superlinear convergence results are proved. It is observed that all the quasi-Newton directions asymptotically approach the Newton direction in both direction and length when the iterate sequence converges to a local minimum of the objective function, and hence the result of superlinear convergence follows. }, issn = {1991-7139}, doi = {https://doi.org/}, url = {http://global-sci.org/intro/article_detail/jcm/8822.html} }
TY - JOUR T1 - Convergence Properties of Multi-Directional Parallel Algorithms for Unconstrained Minimization JO - Journal of Computational Mathematics VL - 4 SP - 357 EP - 372 PY - 2005 DA - 2005/08 SN - 23 DO - http://dor.org/ UR - https://global-sci.org/intro/article_detail/jcm/8822.html KW - Unconstrained minimization KW - Multi-directional parallel quasi-Newton method KW - Global convergece KW - Superlinear convergence AB - Convergence properties of a class of multi-directional parallel quasi-Newton algorithms for the solution of unconstrained minimization problems are studied in this paper. At each iteration these algorithms generate several different quasi-Newton directions, and then apply line searches to determine step lengths along each direction, simultaneously. The next iterate is obtained among these trail points by choosing the lowest point in the sense of function reductions. Different quasi-Newton updating formulas from the Broyden family are used to generate a main sequence of Hessian matrix approximations. Based on the BFGS and the modified BFGS updating formulas, the global and superlinear convergence results are proved. It is observed that all the quasi-Newton directions asymptotically approach the Newton direction in both direction and length when the iterate sequence converges to a local minimum of the objective function, and hence the result of superlinear convergence follows.
Cheng-Xian Xu & Yue-Ting Yang. (1970). Convergence Properties of Multi-Directional Parallel Algorithms for Unconstrained Minimization. Journal of Computational Mathematics. 23 (4). 357-372. doi:
Copy to clipboard
The citation has been copied to your clipboard