Investigation of influence of objective function valley ratio on the determination error of its minimum coordinates
https://doi.org/10.32362/2500-316X-2023-11-6-57-67
Abstract
Objectives. A valley is a region of an objective function landscape in which the function varies along one direction more slowly than along other directions. In order to determine the error of the objective function minimum location in such regions, it is necessary to analyze relations of valley parameters.
Methods. A special test function was used in numerical experiments to model valleys with variables across wide ranges of parameters. The position and other valley parameters were defined randomly. Valley dimensionality and ratio were estimated from eigenvalues of the approximated Hessian of objective function in the termination point of minimum search. The error was defined as the Euclidian distance between the known minimum position and the minimum search termination point. Linear regression analysis and approximation with an artificial neural network model were used for statistical processing of experimental data.
Results. A linear relation of logarithm of valley ratio to logarithm of minimum position error was obtained. Here, the determination coefficient R2 was ~0.88. By additionally taking into account the Euclidian norm of the objective function gradient in the termination point, R2 can be augmented to ~0.95. However, by using the artificial neural network model, an approximation R2 ~ 0.97 was achieved.
Conclusions. The obtained relations may be used for estimating the expected error of extremum coordinates in optimization problems. The described method can be extended to functions having a valley dimensionality of more than one and to other types of hard-to-optimize algorithms regions of objective function landscapes.
About the Author
A. V. Smirnov
MIREA – Russian Technological University
Russian Federation
Competing Interests:
Russian Federation
Alexander V. Smirnov, Cand. Sci. (Eng.), Professor, Department of Telecommunications, Institute of Radio Electronics and Informatics.
Scopus Author ID 56380930700
78, Vernadskogo pr., Moscow, 119454
Competing Interests:
The author declares no conflicts of interest.
References
1. Gill Ph.E., Murray W., Wright M.H. Prakticheskaya optimizatsiya (Practical Optimization). Transl. from Engl. Moscow: Mir; 1985. 509 p. (in Russ.). [Gill Ph.E., Murray W., Wright M.H. Practical Optimization. London: Academic Press; 1981. 402 p.]
2. Ninul A.S. Optimizatsiya tselevykh funktsii. Analitika. Chislennye metody. Planirovanie eksperimenta (Optimization of Objective Functions. Analytics. Numerical Methods. Desing of Experiments). Moscow: Fizmatizdat; 2009. 336 p. (in Russ.).
3. Karpenko A.P. Sovremennye algoritmy poiskovoi optimizatsii. Algoritmy, vdokhnovlennye prirodoi: uchebnoe posobie (Modern Search Optimization Algorithms. Nature-Inspired Optimization Algorithms). Moscow: Izd. MGTU im. Baumana; 2017. 446 p. (in Russ).
4. JasbirS.A. Introduction to optimum design. 4th edition. Elsevier, 2017. 670 p. https://doi.org/10.1016/C2013-0-15344-5
5. Mersmann O., Bischl B., Trautmann H., Preuss M., Weihs C., Rudolf G. Exploratory Landscape Analysis. In: Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation (GECCO’11). 2011. P. 829–836. https://doi.org/10.1145/2001576.2001690
6. Kerschke P., Trautmann H. Comprehensive Feature-Based Landscape Analysis of Continuous and Constrained Optimization Problems Using the R-package flacco. In: Applications in Statistical Computing—From Music Data Analysis to Industrial Quality Improvement. Series: Studies in Classification, Data Analysis, and Knowledge Organization. Bauer N., Ickstadt K., Lübke K., Szepannek G., Trautmann H., Vichi M. (Eds.). Berlin/Heidelberg, Germany: Springer; 2019. P. 93–123. https://doi.org/10.1007/978-3-030-25147-5_7
7. Trajanov R., Dimeski S., Popovski M., Korosec P., Eftimov T. Explainable Landscape-Aware Optimization Performance Predicion. Preprint. 2021. https://arxiv.org/pdf/2110.11633v1
8. Chernorutskii I.G. Metody optimizatsii v teorii upravleniya: uchebnoe posobie (Optimization Methods in the Control Theory). St. Petersburg: Piter; 2004. 256 p. (in Russ.).
9. Smirnov A.V. Properties of objective functions and search algorithms in multi-objective optimization problems. Russ. Technol. J. 2022;10(4):75−85 (in Russ.). https://doi.org/10.32362/2500-316X-2022-10-4-75-85
10. He J., Xu T. New Methods of Studying Valley Fitness Landscapes. Preprint. 2018. https://arxiv.org/pdf/1805.00092.pdf
11. Ochoa G., Veerapen N. Mapping the Global Structure of TSP Fitness Landscapes. J. Heuristics. 2018;24(4):265–294. https://link.springer.com/article/10.1007/s10732-017-9334-0
12. Hansen N., Finck S., Ros R., Auger A. Real-Parameter Black-Box Optimization Benchmarking 2009: Noiseless Functions Definitions. [Research Report] RR-6829, INRIA. 2009. https://hal.inria.fr/inria-00362633v2
13. Jamil M., Yang Xin-She. A literature survey of benchmark functions for global optimization problems. Int. J. Math. Model. Numer. Opt. 2013;4(2):150–194. https://doi.org/10.1504/IJMMNO.2013.055204
14. Bol’shakov A.A., Karimov R.N. Metody obrabotki mnogomernykh dannykh i vremennykh ryadov (Methods for Processing of Multidimensional Data and Time Series). Moscow; Goryachaya Liniya–Telekom; 2007. 522 p. (in Russ.).
15. Haykin S. Neironnye seti: polnyi kurs (Neural networks. A Comprehensive Foundation). Transl. from Engl. Moscow: Vil’yams; 2006. 1104 p. (in Russ.). [Haykin S. Neural networks. A Comprehensive Foundation. Prentice Hall; 1999. 1104 p.]
Supplementary files
|
|
1. Histograms of the main experiment results | |
| Subject | ||
| Type | Исследовательские инструменты | |
View
(35KB)
|
Indexing metadata ▾ | |
- A valley is a region of an objective function landscape in which the function varies along one direction more slowly than along other directions.
- A linear relation of logarithm of valley ratio to logarithm of minimum position error was obtained.
- The obtained relations may be used for estimating the expected error of extremum coordinates in optimization problems.
Review
For citations:
Smirnov A.V. Investigation of influence of objective function valley ratio on the determination error of its minimum coordinates. Russian Technological Journal. 2023;11(6):57-67. https://doi.org/10.32362/2500-316X-2023-11-6-57-67
Views:
300
Email this article 