Conclusion
After the burst of enthusiasm caused by the Khachiyan result [5], the methods of ellipsoids turned out to be the center of attention of many optimizers. After some time, only their theoretical value began to be achnowledged, raising doubts about their practical efficiency. And this is valid to an extent if we speak of the MCGM method and its negligible modifications.
In this paper the authors have attempted to show that ellipsoid methods are just a very particular case of the families of gradient type algorithms using a space stretching operation. Other representatives of this family, as r-algorithms, say, are an effective practical means for solving many complex mathematical programming problems reducing to nondifferentiable optimization. The theory of a whole class of algorithms with space stretching is still far from completion. It seems to us a sufficiently realistic aim to construct, such an algorithm, which would be no less practically efficient than the r-algorithm and would have as good a foundation as the method of ellipsoids.
Similar content being viewed by others
Literature Cited
V. I. Gershovich, “On a truncation method using linear transformation of space,” in: Theory of Optimal Solutions [in Russian], Inst. Kibern. Akad. Nauk Ukr. SSR, Kiev (1979). pp. 15–23.
V. I. Gershovich, “On a minimization method using linear transformation of space,” in: Theory of Optimal Solutions [in Russian], Inst. Kibern. Akad. Nauk Ukr. SSR, Kiev (1980), pp. 38–45.
V. I. Gershovich, “On an algorithm of ellipsoids,” in: On Certain Algorithms of Nonsmooth Optimization and Discrete Programming [in Russian], Kibern. Akad. Nauk Ukr. SSR, Kiev (1981), pp. 8–13.
A. Yu. Levin, “On a minimization algorithm for convex functions,” Dokl. Akad. Nauk SSSR,160, No. 6, 1244–1247 (1965).
L. G. Khachiyan, “Polynomial algorithm in linear programming,” Dokl. Akad. Nauk SSSR,244, No. 5, 1093–1096 (1979).
L. G. Khachiyan, “Polynomial algorithm in linear programming,” Vychisl. Mat. Mat. Fiz.,20, No. 2, 51–68 (1979).
N. Z. Shor, “Use of the space stretching operations in convex function minimization problems,” Kibernetika, No. 1, 2–12 (1970).
N. Z. Shor “On the convergence rate for the method of generalized gradient descent with space stretching,” Kibernetika, No. 2, 80–85 (1970).
N. Z. Shor and N. G. Zhurbenko, “Minimization method using the space stretching operation in the direction of the difference between two successive gradients,” Kibernetika, No. 3, 51–59 (1971).
N. Z. Shor, “On the method of minimizing almost-differentiable functions,” Kibernetika, No. 4, 65–70 (1972).
N. Z. Shor, “Truncation method with space stretching for the solution of convex programming problems,” Kibernetika, No. 1, 94–95 (1977).
N. Z. Shor, Methods of Minimizing Nondifferentiable Functions and Their Application [in Russian], Naukova Dumka, Kiev (1979).
N. Z. Shor and V. I. Gershovich, “On a family of algorithms to solve convex programming problems,” Kibernetika, No. 4, 62–67 (1979).
D. B. Yudin and A. S. Nemirovskii, “Information complexity and effective methods of solving convex extremal problems,” Ekon. Mat. Metody,12, No. 2, 357–369 (1979).
M. K. Kozlov, S. P. Tarasov, and L. G. Khakiyan, “Polynomial solvability of convex quadratic programming,” Dokl. Akad. Nauk SSSR,248, No. 5, 1049–1051 (1979).
N. Z. Shor and V. I. Gershovich, “On a modification of gradient type algorithms with space stretching to solve large dimensionality problems,” Kibernetika, No. 5, 67–70 (1981).
P. Wolfe, A Bibliography for the Ellipsoid Algorithm, RC 8237, IBM Research Center, New York (1980).
A. Bachem and M. Grotschel, “Characterizations of adjacency of faces of polyhedra,” Mat. Progr. Study, No. 14, 1–22 (1981).
R. Shrader, “Ellipsoid methods,” Inst. for Econometry and Operations Research, No. 81174-OR Bonn (1980).
M. Grotschel, L. Lovasz, and A. Schriyver, “The ellipsoid method and its consequences in combinatorial optimization,” Combinatorica,1, No. 2, 169–197 (1981).
B. Korte and R. Shrader, “A note on convergence proofs for Shor-Khachiyan methods,” Lect. Notes Control and Information Sci., No. 30, 51–57 (1980).
L. Lovasz, “On the Shannon capacity of a graph,” IEEE Trans. Inf. Theory,25, No. 1, 1–7.
M. Grotschel, L. Lovasz, and A. Schrijver, “Polynomial algorithms for perfect graphs,” Inst. for Econometry and Operations Research, WP 81176-OR, Bonn (1981).
Additional information
Translated from Kibernetika, No. 5, pp. 61–69, September–October, 1982.
Rights and permissions
About this article
Cite this article
Gershovich, V.I., Shor, N.Z. Method of ellipsoids, its generalizations and applications. Cybern Syst Anal 18, 606–617 (1982). https://doi.org/10.1007/BF01068741
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01068741