Eigenvalue Decomposition Based Modified Newton Algorithm

Wen Jun Wang; Ju Bo Zhu; Xiao Jun  Duan

doi:10.4028/www.scientific.net/AMM.347-350.2586

Paper Titles

The Petri Net Model of the Requirements and Service Composition
p.2564

The Design and Implementation of Florists Shop Online System Based on J2EE
p.2571

Web Page Data Collection Based on Multithread
p.2575

Functional Link Artificial Neural Networks Filter for Gaussian Noise
p.2580

Eigenvalue Decomposition Based Modified Newton Algorithm
p.2586

Bayesian Networks Application in Multi-State System Reliability Analysis
p.2590

Noether Theorem of a Kind of Singular Integral Equation with Hilbert Kernel on Closed Contours
p.2596

Adaptive Dual Conjudate Gradient Projection Algorithm for Compressed Sensing Image Reconstruction
p.2600

A Novel Model-Based Method for Automatic Generation of FMEA
p.2605

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 347-350Eigenvalue Decomposition Based Modified Newton...

Eigenvalue Decomposition Based Modified Newton Algorithm

Abstract:

When the Hessian matrix is not positive, the Newton direction maybe not the descending direction. A new method named eigenvalue decomposition based modified Newton algorithm is presented, which first takes eigenvalue decomposition on the Hessian matrix, then replaces the negative eigenvalues with their absolutely values, finally reconstruct Hessian matrix and modify searching direction. The new searching direction is always the descending direction, and the convergence of the algorithm is proved and conclusion on convergence rate is presented qualitatively. At last, a numerical experiment is given for comparing the convergence domains of modified algorithm and classical algorithm.

You might also be interested in these eBooks

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volumes 347-350)

Pages:

2586-2589

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.347-350.2586

Citation:

Cite this paper

Online since:

August 2013

Authors:

Wen Jun Wang, Ju Bo Zhu, Xiao Jun Duan

Keywords:

Convergence, Convergence Domain, Eigenvalue Decomposition, Newton Algorithm

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

References

[1] D. P. Bertsekas. Nonlinear Programming. Athena Scientific, Belmont, Massachussetts, (1999).

Google Scholar

[2] S. Boyd and L. Vandenberghe, Convex optimization, Cambridge University Press, Cambridge, UK, (2004).

Google Scholar

[3] A. Goldstein and J. F. Price, On Descent from Local Minima, Mathematics of Computation, Volume 25, No. 115, July, 1971, pp: 569-574.

DOI: 10.1090/s0025-5718-1971-0312365-x

Google Scholar

[4] AV. Fiacco, GP. McCormick, The sequential unconstrained minimization technique for nonlinear-a primal-dual method, Management Science, Vol 10, No. 2, 1964, pp: 360-365.

DOI: 10.1287/mnsc.10.2.360

Google Scholar

[5] J.M. Ortega, W.C. Rheinboldt, Iterative Solution of Nonlinear Equations in Several Variables, Academic Press, New York-London, (1970).

DOI: 10.1016/b978-0-12-528550-6.50017-x

Google Scholar

[6] R. Horn and C. R. Johnson, Matrix analysis, Cambridge University, Press, New York, (1985).

Google Scholar

[7] M. J. D. Powell, An e_cient method for _nding the minimum of a function of several variables without calculating derivatives, The Computer Journal, 7, pp.155-162, (1964).

DOI: 10.1093/comjnl/7.2.155

Google Scholar