Summary
We present an algorithm which enables us to calculate one particular subgradient of a convex functionf: ℝ2→ℝ at a given point. Such a calculation is required in many existing numerical methods for convex nondifferentiable optimization. The novelty of our approach lies in the assumption that only the values off are computable and no analytical formula for the subdifferential is known. We include some numerical examples.
Similar content being viewed by others
References
Gaudioso, M.: An algorithm for convex NDO based on properties of the contour lines of convex quadratic functions. In: Demyanov, V.F., Pallaschke, D. (eds.) Nondifferentiable optimization: Motivations and applications pp. 190–196. Lect. Notes Econ. Math. Syst., vol. 255. Berlin, Heidelberg, New York, Tokyo: Springer 1985
Demyanov, V.F., Vasilev, L.V.: Nondifferentiable optimization. New York: Optimization Software Berlin, Heidelberg, New York, Tokyo: Springer 1985. Russian edition: Moscow: Nauka 1981
Kiwiel, K.C.: An algorithm for linearly constrained convex nondifferentiable minimization problems. J. Math. Anal. Appl.105, 452–465 (1985)
Kiwiel, K.C.: An exact penalty function algorithm for non-smooth convex constrained minimization problems. IMA J. Numer. Anal.5, 111–119 (1985)
Kiwiel, K.C.: A constraint linearization method for nondifferentiable convex minimization. Numer. Math.51, 395–414 (1987)
Lemaréchal, C.: Constructing bundle methods for convex optimization. In: Hiriart-Urruty, J.-B. (ed.) Fermat Days 85: Mathematics for Optimization pp. 201–240. Math. Stud., vol. 129. Amsterdam: North-Holland 1986
Rockafellar, R.T.: Convex analysis. Princeton, New Jersey: Princeton University Press 1970
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Studniarski, M. An algorithm for calculating one subgradient of a convex function of two variables. Numer. Math. 55, 685–693 (1989). https://doi.org/10.1007/BF01389336
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01389336