IJPAM: Volume 63, No. 2 (2010)


Melisa Hendrata$^1$, P.K. Subramanian$^2$
$^{1,2}$Department of Mathematics
College of Natural and Social Sciences
California State University
5151, Los Angeles, State University Drive, Los Angeles, CA 90032-8103, USA
$^1$e-mail: [email protected]
$^2$e-mail: [email protected]

Abstract.Newton's algorithm and some of its variations are often used to find global minima of real valued functions. We propose another such variation and prove its local quadratic convergence. We combine this with Armijo type line search method to produce global convergence, which is eventually quadratic, for the important class of strictly convex functions. Computational performance on some standard test problems is presented, which shows that the proposed model may be viable.

Received: July 12, 2010

AMS Subject Classification: 90xxx, 90-08

Key Words and Phrases: Newton's method, Armijo line search, global optimization

Source: International Journal of Pure and Applied Mathematics
ISSN: 1311-8080
Year: 2010
Volume: 63
Issue: 2