Abstract
We consider quasidifferentiable functions in the sense of Demyanov and Rubinov, i. e. functions, which are directionally differentiable and whose directional derivative can be expressed as a difference of two sublinear functions, so that its “subdifferential”, called the quasidifferential, consists of a pair of sets. For these functions a generalized gradient algorithm is proposed. Its behaviour is studied in detail for the special class of continuously subdifferentiable functions. Numerical test results are given. Finally, the general quasidifferentiable case is simulated by means of “perturbed” subdifferentials, where we make use of the non-uniqueness in the quasidifferential representation.
Similar content being viewed by others
References
A.M. Bagirov, “Numerical methods for minimizing quasidifferentiable functions: A survey and comparison,” in Quasidifferentiability and Related Topics, Ser.: Nonconvex Optimization and its Applications, Demyanov V.F. and Rubinov A.M. (Eds.), vol. 43, pp. 33–71, Kluwer Academic Publishers: Dordrecht, 2000.
D.P. Bertsekas, Constrained Optimization and Lagrange Multiplier Methods, Academic Press: New York, 1982.
V.F. Demyanov, S. Gamidov, and T.I. Sivelina, “An algorithm for minimizing a certain class of quasidifferentiable functions,” Math. Programming Study, vol. 29, pp. 74–84, 1986.
V.F. Demyanov and A.M. Rubinov, Quasidifferential Calculus, Optimization Software, Inc., Publications Division: New York, 1986.
H. Jandl, On the Implementation of Quasidifferentiable Optimization Problems Under Constraints, Fortschritt-5. H. Jandl, On the Implementation of Quasidifferentiable Optimization Problems Under Constraints, Fortschritt-Berichte VDI, 16/44, VDI-Verlag, Duesseldorf, 1988.
L. Kuntz, “A characterization of continuously codifferentiable functions and some consequences,” Optimization, vol. 22, pp. 539–547, 1991.
L. Kuntz and A. Pielczyk, “The method of common descent for a certain class of quasidifferentiable functions,” Optimization, vol. 22, pp. 669–679, 1991.
B. Luderer and J. Weigelt, “A generalized steepest descent method for continuously subdifferentiable functions,” Optimization, vol. 30, pp. 119–135, 1994.
D. Pallaschke and P. Recht, “On the steepest-descent method for a class of quasidifferentiable optimization problems,” in Nondifferentiable Optimization: Motivations and Applications, Proc. IIASAWorkshop (Sopron, 1984), Lect. Notes Econ. Math. Syst., vol. 255, Springer-Verlag: Berlin 1985, pp. 252–263.
H. Schramm, Eine Kombination von Bundle-und Trust-Region-Verfahren zur Loesung nichtdifferenzierbarer Optimierungsprobleme, Bayreuth. Math. Schr., vol. 30, 1989.
L.W. Zhang and Z.Q. Xia, “Newton-type methods for quasidifferential equations,” J. Optimization Theory Appl., vol. 108, no. 2, pp. 439–456, 2001.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Luderer, B., Weigelt, J. A Solution Method for a Special Class of Nondifferentiable Unconstrained Optimization Problems. Computational Optimization and Applications 24, 83–93 (2003). https://doi.org/10.1023/A:1021854116784
Issue Date:
DOI: https://doi.org/10.1023/A:1021854116784