•  

Separation of parameters using machine learning methods in global optimization problems

TitleSeparation of parameters using machine learning methods in global optimization problems
AuthorsK. A. Barkalov1, M. A. Usova1
1Lobachevsky State University of Nizhny Novgorod Nizhny Novgorod, Russia
AnnotationThe paper presents the findings from research into an approach to solving global optimization problems with a different nature of dependence on diverse groups of parameters. A scheme for choosing the problem parameters, which have a local effect on the objective function is proposed, which allows to solve essentially multidimensional problems using the nested optimization scheme. At the same time, different optimization algorithms are used at differing levels of recursion (nesting levels). The proposed approach has demonstrated its efficiency in solving several series of test problems. The effectiveness of the proposed approach was studied.
Keywordsglobal optimization, local optimization, nested optimization scheme, highdimensional problems, parameter separation
CitationBarkalov K. A., Usova M. A. ''Separation of parameters using machine learning methods in global optimization problems'' [Electronic resource]. Proceedings of the International Scientific Youth School-Seminar "Mathematical Modeling, Numerical Methods and Software complexes" named after E.V. Voskresensky (Saransk, July 14-18, 2022). Saransk: SVMO Publ, 2022. - pp. 32-39. Available at: https://conf.svmo.ru/files/2022/papers/paper06.pdf. - Date of access: 30.11.2022.