On the Convergence of Parallel Broyden Method

Gan Zhou Wu

doi:10.4028/www.scientific.net/AMR.143-144.1148

Paper Titles

The Study of the Factors to Establishing an Online Environment for Negotiation
p.1127

Partheno Genetic Ant Colony Optimization Algorithm and its Application
p.1132

Prediction for the Sulfur Content in Pig Iron of Blast Furnace by Combining Artificial Neural Network with Genetic Algorithm
p.1137

Four Quadrants Numerical Prediction for Hydrodynamic Performance of Open-Water Propeller
p.1143

On the Convergence of Parallel Broyden Method
p.1148

An Improved PSO-ACO Algorithm for Solving Large-Scale TSP
p.1154

OneAPI Restful Services and Java Implementation
p.1159

The Mid-Long Term Power Load Forecasting Based on Gray and SVM Algorithm
p.1164

The Converse Theorem of Lyapunov Exponential Stability Theorem
p.1170

HomeAdvanced Materials ResearchAdvanced Materials Research Vols. 143-144On the Convergence of Parallel Broyden Method

On the Convergence of Parallel Broyden Method

Abstract:

Compare to the sequential Broyden method, an asynchronous parallel Broyden method is presented in the paper. We suppose that we have processors, which are divided into two groups. And the first group has processors, the second processors has processors, while the two groups are asynchronous parallel. If we assume that the objective function is twice continuously differentiable and uniformly convex, the global convergence of the algorithm is given. And under the same conditions, we show that the parallel Broyden method is superlinearly convergent.

You might also be interested in these eBooks

Smart Materials and Intelligent Systems, SMIS2010

View Preview

Info:

Periodical:

Advanced Materials Research (Volumes 143-144)

Pages:

1148-1153

DOI:

https://doi.org/10.4028/www.scientific.net/AMR.143-144.1148

Citation:

Cite this paper

Online since:

October 2010

Authors:

Gan Zhou Wu

Keywords:

Global Convergence, Parallel Broyden Method, Superlinear Convergence, Unconstrained Optimization Problem

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

References

[1] Fisher H, Ritter k. An asynchronous parallel Newton methods. Math Prog, 1988, 42: 363-374.

Google Scholar

[2] Conforti D, Musmanno R. A parallel asynchronous Newton algorithm for unconstrained optimization. JOTA, 1993, 77: 305-322.

DOI: 10.1007/bf00940714

Google Scholar

[3] Felgenhauer U. On the stable convergence of particular quasi-Newton methods. Optimization, 1992, 26: 97-117.

DOI: 10.1080/02331939208843845

Google Scholar

[4] .

Google Scholar

[4] Byrd R H, Nocedal J. A tool for the analysis of quasi-Newton methods with application to unconstrained minimization. SIAM J. Numer. Anal, 1989, 26: 727-739.

DOI: 10.1137/0726042

Google Scholar

[5] Powell M J D. Some global convergence properties of a variable metric algorithm for minimization without exact line searches. In : cottle R W eds. Nonlinear Programming, SIAM-AMS Proceedings. Americal Mathematical Society, Providence, RI, (1976).

Google Scholar

[6] Pearson J D. Variable metric methods of minimization. Computer J, 1969, 12: 171-178.

Google Scholar

[7] Byrd R H, Nocedal J, Yuan Y X. Global convergence of quasi-Newton methods on convex problems. SLAM J, Num, Anal, 1987, 24： 1171-1190.

DOI: 10.1137/0724077

Google Scholar

[8] Griewank A, and Toint P L, Local convergence analysis for partitioned quasi-linear updates, Numer. Math, 1982, 39: 429-448.

DOI: 10.1007/bf01407874

Google Scholar

[9] Dennis J E, More J, A characterization of superlinear convergence and its application to quasi-Newton methods. Math computer. 1974, 28: 549-560.

DOI: 10.1090/s0025-5718-1974-0343581-1

Google Scholar