Abstract
The paper deals with a numerical minimization problem for a convex function defined on a convexn-dimensional domain and continuous (but not necessarily smooth). The values of the function can be calculated at any given point. It is required to find the minimum with desired accuracy. A new algorithm for solving this problem is presented, whose computational complexity asn → ∞ is considerably less than that of similar algorithms known to the author. In fact, the complexity is improved fromCn 7 ln2(n+1) [4] toCn 2 ln(n+1).
Similar content being viewed by others
References
A. S. Nemirovskii and D. B. Yudin, “Informative complexity and effective methods for solving convex extremal problems,”Ékonom. i Mat. Metody,12, No. 2, 357–369 (1976).
A. Yu. Levin, “An algorithm for minimizing convex functions,”Dokl. Akad. Nauk SSSR [Soviet Math. Dokl.],160, No. 6, 1241–1243 (1965).
D. B. Yudin, D. M. Goryashko, and A. S. Nemirovskii,Mathematical Methods for Optimizing Devices and Algorithms Used in Automatic Control Systems [in Russian], Radio i Svyaz', Moscow (1982).
M. de Guzman,Differentiation of Integrals in ℝn, Springer-Verlag, Berlin-Heidelberg-New York (1975).
B. S. Mityagin, “Two inequalities for volumes of convex bodies,”Mat. Zametki [Math. Notes],5, No. 5, 20–26 (1959).
S. P. Tarasov, L. G. Kachiyan, and I. I. Erlikh, “Insrcibed ellipsoid method,”Dokl. Akad. Nauk SSSR [Soviet Math. Dokl.],298, No. 5 (1988).
Author information
Authors and Affiliations
Additional information
Translated fromMatematicheskie Zametki, Vol. 59, No. 1, pp. 95–102, January, 1996.
Rights and permissions
About this article
Cite this article
Protasov, V.Y. Algorithms for approximate calculation of the minimum of a convex function from its values. Math Notes 59, 69–74 (1996). https://doi.org/10.1007/BF02312467
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02312467