Abstract
In this paper, we present an asynchronous parallel quasi-Newton method. We assume that we havep+q processors, which are divided into two groups, the two groups execute in an asynchronous parallel fashion. If we assume that the objective function is twice continuously differentiable and uniformly convex, we discuss the global and superlinear convergence of the parallel BFGS method. Finally we show numerical results of this algorithm.
Zusammenfassung
Dieser Beitrag stellt ein asynchron paralleles Quasi-Newton-Verfahren vor. Wir setzen voraus, daß esp+q Prozessoren gibt, die in zwei Gruppen geteilt sind. Die beiden Gruppen arbeiten asynchron parallel. Wenn die Zielfunktion zweimal stetig differenzierbar und gleichmäßig konvex ist, beweisen wir, daß die vom Verfahren erzeugte Iterationsfolge zur Optimallösung global und superlinear konvergiert.
Similar content being viewed by others
References
Byrd, R. H., Nocedal, J.: A tool for the analysis of quasi-Newton methods with application to unconstrained minimization. SIAM J. Numer. Anal.26, 727–739, (1989).
Byrd, R. H., Nocedal, J., Yuan, Y.: Global convergence of a class of quasi-Newton methods on convex problems. SIAM J. Numer. Anal.24, 1171–1189 (1989).
Dennis, J. E., Moré, J. J.: A characterization of superlinear convergence and its application to quasi-Newton methods. Math. Comput.28, 549–560 (1974).
Pardalos, P. M., Phillips, A. T., Rosen, J. B.: Topics in parallel computing in mathematical programming. New York: Science Press 1993.
Griewank, A., Toint, Ph L.: Local convergence analysis for partitioned quasi-Newton updates. Numer. Math.39, 429–448 (1982).
Pearson, J. D.: Variable metric methods of minimization. Comput. J.12, 171–178 (1969).
Powell, M. J. D.: Some global convergence properties of a variable metric algorithm for minimization without exact line searches. In: Nonlinear programming, SIAM-AMS Proceedings, vol. 9 (Cottle, R. W., eds.). Americal Mathematical Society, Providence, RI, 1976.
Chen, Z. Fei, P.: A parallel algorithm for constrained optimization problems. J. Comp. Appl. Math.61 (1995).
Author information
Authors and Affiliations
Additional information
Supported by NNSF of China and National 863 HTP.
Rights and permissions
About this article
Cite this article
Chen, Z., Fei, P. & Zheng, H. A parallel quasi-Newton algorithm for unconstrained optimization. Computing 55, 125–133 (1995). https://doi.org/10.1007/BF02238097
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02238097