Abstract
We consider an algorithm with space dilation. For a certain choice of the dilation coefficient, this is a method of outer approximation of semi-ellipsoids by ellipsoids with monotonous decrease in their volume. It is shown that the Yudin-Nemirovski-Shor ellipsoid method is a specific case. Two forms of the algorithm are dealt with: the B-form, where the inverse space transformation matrix B is updated, and the H-form, where the symmetric matrix \(H=BB^{\top }\) is updated. Our test results show that the B-form of the algorithm is computationally more robust to error accumulation than the H-form. The application of the algorithm for finding a minimizer of a convex function, for solving convex programming problems, and for determining a saddle point of a convex-concave function is described. Possible ways of accelerating the algorithm by deeper ellipsoid approximations are discussed as well.
Supported by Volkswagen Foundation (grant No 90 306).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Yudin, D.B., Nemirovskii, A.S.: Informational complexity and efficient methods for the solution of convex extremal problems. Matekon 13(3), 25–45 (1976)
Shor, N.Z.: Cut-off method with space dilation in convex programming problems. Cybernetics 13(1), 94–96 (1977)
Stetsyuk, P.I., Fesiuk, O.V., Khomyak, O.N.: The generalized ellipsoid method. Cybern. Syst. Anal. 54(4), 576–584 (2018)
Stetsyuk, P.I.: An approximate method of ellipsoids. Cybern. Syst. Anal. 39(3), 435–439 (2003)
Stetsyuk, P.I., Fischer, A., Lyashko, V.I.: An ellipsoid method for minimization of convex function. NaUKMA Res. Pap. Comput. Sci. 2, 16–21 (2019). https://doi.org/10.18523/2617-3808.2019.2.16-21. (in Ukranian)
Stetsyuk, P.I.: 2D-ellipsoid of optimal volume and its applications. In: Polyakova, L.N. (ed.) Constructive Nonsmooth Analysis and Related Topics (Dedicated to the Memory of V.F. Demyanov) (CNSA), pp. 303–306. IEEE (2017)
Shor, N.Z.: Minimization Methods for Non-Differentiable Functions. Springer, Heidelberg (1985). https://doi.org/10.1007/978-3-642-82118-9
Stetsyuk, P.I.: Shor’s r-algorithms: theory and practice. In: Butenko, S., Pardalos, P.M., Shylo, V. (eds.) Optimization Methods and Applications. SOIA, vol. 130, pp. 495–520. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-68640-0_24
Stetsyuk, P.I.: Theory and software implementations of Shor’s r-algorithms. Cybern. Syst. Anal. 53(4), 692–703 (2017)
Stetsyuk, P.I., Fischer, A.: Shor’s r-algorithms and octave-function ralgb5a. In: International Conference “Modern Informatics: Problems, Achievements and Prospects for Development”, Ukraine, Kyiv, December 2017. V.M. Glushkov Institute of Cybernetics of the NAS of Ukraine, pp. 143–146 (2017). (in Russian)
Sergienko, I.V., Stetsyuk, P.I.: On N.Z. Shor’s three scientific ideas. Cybern. Syst. Anal. 48(1), 2–16 (2012)
Stetsyuk, P.I.: \(r\)-Algorithms and ellipsoids. Cybern. Syst. Anal. 32(1), 93–110 (1996)
Stetsyuk, P.I.: Methods of ellipsoids and r-algorithms, Eureka, Chisinau, Moldova (2014). (in Russian)
Acknowledgement
The authors would like to thank Oleksii Lykhovyd and Volodymyr Zhydkov for their help in preparing this paper. Moreover, the authors are grateful to an anonymous referee for comments that helped to improve the paper.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Stetsyuk, P., Fischer, A., Khomyak, O. (2020). The Generalized Ellipsoid Method and Its Implementation. In: Jaćimović, M., Khachay, M., Malkova, V., Posypkin, M. (eds) Optimization and Applications. OPTIMA 2019. Communications in Computer and Information Science, vol 1145. Springer, Cham. https://doi.org/10.1007/978-3-030-38603-0_26
Download citation
DOI: https://doi.org/10.1007/978-3-030-38603-0_26
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-38602-3
Online ISBN: 978-3-030-38603-0
eBook Packages: Computer ScienceComputer Science (R0)