XXV International conference “Mathematical Optimization Theory and Operations Research” (MOTOR 2026)

Invited Speakers

 

Invited Speakers

 

	Prof. Saeid Alikhani Department of Mathematics, Yazd University, Yazd, Iran Majority Distinguishing Colorings of Graphs We study graph colorings that simultaneously satisfy two constraints: the majority condition and the distinguishing property. A majority coloring requires that, at each vertex, no color appears on more than half of the adjacent vertices or incident edges, while a distinguishing coloring breaks all nontrivial automorphisms of the graph. Motivated by recent work on majority distinguishing edge colorings, we introduce and formalize the concept of majority distinguishing vertex colorings. Accordingly, we define and study two new graph parameters, called the majority distinguishing number and the majority distinguishing index. We determine exact values of these parameters for several fundamental graph families, including paths, cycles, friendship graphs, book graphs, and certain classes of trees. In addition, we investigate how these parameters behave under the corona product of graphs and establish general bounds and relationships with classical distinguishing parameters. Our results extend and unify previous studies on symmetry breaking and majority colorings, and they suggest several open problems and directions for further research.
	Prof. Mikhail Y. Kovalyov United Institute of Informatics Problems, National Academy of Sciences of Belarus, Minsk, Belarus Routing and Charging Scheduling for Regular Electric Buses This work is inspired by the results of the project ‘‘Planning Process and Tool for Step-by-Step Conversion of the Conventional or Mixed Bus Fleet to a 100% Electric Bus Fleet’’ of the Electric Mobility Europe initiative. The public transport operators involved in this project faced the challenge of effectively routing and charging electric buses operating on a fixed timetable. They have all the data to make appropriate decisions. Currently, decisions are made primarily based on previous routing schemes and the "first come, first charged" principle.  The number of chargers at a charging station is determined by the maximum number of electric buses that can simultaneously reside at the station between trips. We provide an overview of research in the field of routing and charging scheduling for urban electric buses. We discuss the relationship between these problems and classic vehicle routing and machine scheduling problems.
	Prof. Huỳnh Thị Thanh Bình School of Information and Communication Technology (SoICT), Hanoi University of Science and Technology (HUST), Hanoi, Vietnam Evolutionary Multitasking: Recent Advances and Applications Evolutionary multitasking optimization is a cutting-edge topic in the field of computational intelligence that merges evolutionary computation and multitasking methodologies to address multiple optimization problems concurrently. By leveraging the interactions and dependencies between problems, evolutionary multitasking allows for the sharing and transfer of valuable information between tasks, thereby enhancing the overall search performance. This talk will provide an overview of the fundamental concepts, principles, and recent advancements of evolutionary multitasking optimization. Additionally, the talk will highlight the practical significance of evolutionary multitasking algorithms and demonstrate how these techniques can effectively tackle complex real-world optimization problems. Through these problems, the talk will delve into algorithm design issues such as solution encoding and the knowledge transfer mechanism in the multitasking environment. Performance evaluations on benchmark datasets will also be presented to demonstrate the effectiveness of evolutionary multitasking algorithms.
	Prof. Mikhail A. Marchenko Artificial Intelligence and Information Technology Laboratory, Institute of Computational Mathematics and Mathematical Geophysics SB RAS, Novosibirsk, Russia Parallelization of statistical modeling methods The talk will focus on methods of parallelization of statistical modeling algorithms. The main emphasis is placed on the problems of statistical modeling of kinetic processes, what is due to their practical significance. An important area touched upon in the talk is the optimization of parallel statistical modeling algorithms in order to reduce their computational complexity. The level of performance of modern supercomputers makes the use of probabilistic models of kinetic processes extremely relevant. First reason is that such models adequately describe physical phenomena considering the influence of rare events, which is practically impossible for other approaches. Next reason is that they can be effectively implemented as parallel programs. In numerical statistical modeling applications, it is necessary to use parallel pseudorandom number generators, effectively implemented on multiprocessor computers. Then it is necessary to apply a distributed computing methodology that allows for parametric analysis of probabilistic models. The last problem is related to the study of the characteristics of probabilistic models, such as estimation error and computational labor intensity, depending on the parameters of the problem and the algorithm. To solve these problems, it is also necessary to have standard parallelized software tools.
	Prof. Elena V. Konstantinova Sobolev Institute of Mathematics SB RAS, Novosibirsk, Russia CayleyPy project: artificial intelligence methods for Cayley graphs In this talk we review some new results obtained by using AI methods to check open problems in Cayley graph theory and to state new conjectures. The results discussed in the talk are published in [1]. This is the third paper of the CayleyPy project, https://github.com/CayleyPy/CayleyPy, applying artificial intelligence methods to problems in group theory. This is the first paper where the public release of CayleyPy is announced. CayleyPy is an open–source Python library for computations with Cayley and Schreier graphs. Compared with state-of-the-art systems based on classical methods, such as GAP and Sage, CayleyPy handles significantly larger graphs and performs several orders of magnitude faster. For many Cayley graphs over the symmetric group Symn quasi–polynomial diameter formulas are observed: a small set of quadratic or linear polynomials on n, and it is conjectured that it is a general phenomenon. These lead to efficient diameter computation, despite the problem being NP-hard in general. A refinement of the Babai-type conjecture on diameters of Symn is proposed as follows: 1/2n2 + 4n upper bounds for the diameters in the standard undirected case, as compared to prior conjectural bounds of O(n2). Explicit generator families, related to involutions in a simple “square-with-whiskers” pattern, are conjectured to maximize the diameter; extensive search confirms this for all n ⩽ 15. Some of conjectures are “LLM-friendly” — they can be stated as sorting problems, which are easy to formulate for LLM, and their solutions can be given by an algorithm or by a Python code, which is easy to verify, so they can be used to test LLM’s abilities to solve research problems. To benchmark various methods of path-finding on Cayley graphs more than 10 benchmark datasets were created in the form of Kaggle challenges, making benchmarking easy and public to the community. CayleyPy works with arbitrary permutation or matrix groups, and supports a pre-defined collection of more than a hundred generators including puzzle groups. The code for direct growth computation outperforms similar functions on the standard computer algebra system GAP/SAGE up to 1000 times both in speed and in maximum sizes of the graphs that it can handle.   [1] A. Chervov, D. Fedoriaka, E. Konstantinova, and et al., CayleyPy Growth: Efficient growth computations and hundreds of new conjectures on Cayley graphs (Brief version), arXiv, 2025, https://doi.org/10.48550/arXiv.2509.19162.

 

Tutorials

 

	Prof. Alexander Strekalovskiy Matrosov Institute for System Dynamics and Control Theory SB RAS, Irkutsk, Russia New Principles of Nonconvex Optimization We address the nonconvex optimization problem with the cost function, equality and inequality constraints given by DC functions (difference of two convex functions).  With the help of the Exact Penalty Theory, one reduces the problem with constraints to an unconstrained (penalized) problem with a DC cost function. A Global Search Scheme (GSS) for solving resulting nonconvex problem combines (a) Special local search methods (LSM) providing a critical (regarding a LSM) point. Within a LSM one can use classical optimization methods and modern computational software, that looks as an advantage of the developed approach; (b) Global Search procedures, suggested by the new approximate Global Optimality Conditions (GOC), and, due to the constructive (algorithmic) property of the GOC, allowing to escape critical, local and stationary points with improvement of the goal function. Hence, a convergency proof of the GSS is not only important for substantiation of the methodology, but crucial for solving a big number of nonconvex problems arising in various applied fields. As well-known, optimization problems can be divided into two main types: convex and non-convex. First one can be solved with classical numerical techniques as Gradient methods, Newtonian ones, IPM, SQP, TRM among others. Besides, modern software (CPLEX, Gurobi, etc.) is highly effective for solving these problems without “curse of dimensionality”. Meanwhile, for a huge area of non-convex problems, convex optimization methods turn out to be completely inefficient and extremely unworkable when they come to find a global solution to problems in question. This makes some simplifying ideas very popular, such as B& B, bio-inspired methods, genetic algorithms and neural networks etc., while these approaches have the lack of any mathematical justifications and background. Numerical testing on high-dimensional test problems certified the comparative effectiveness of the developed approach.
	Prof. Anton Eremeev Omsk Department of Sobolev Institute of Mathematics SB RAS, Omsk, Russia Expected runtime analysis of evolutionary algorithms The theory of runtime analysis of the evolutionary algorithms, including genetic algorithms, has been developed during the last three decades, starting with the works of Ingo Wegener and his co-authors. This theory allows to obtain provable mathematical results that give some guarantees on the expected computational cost till first hitting an optimal solution by the EA population (expected runtime), or to show the inefficiency of some EAs for some parameter settings. This tutorial surveys some recent techniques that proved to be useful in runtime analysis of EAs and discusses the open problems in the area.
	Prof. Alexey Chernov LLC «Quantum Algorithms And Optimization» Dolgoprudny, Russia The universal solver QORETEX (Orsima). What has changed over the past year The report will briefly describe the innovations in the solver over the past year, the successes and difficulties that arose. In particular, the following topics will be covered:  - Creation of a library of linear algebra, its features for working with large dimensions and its adaptation to the linear solver.  - Development of the problem preprocessing module, including a description of some presolving rules, the total number of which reaches 50. This already opens up the possibility for us to solve problems inaccessible to other solvers.  - Implemented cuts and heuristics for finding integer points.  - Used methods for solving linear programming problems and some features of their implementation. An important factor for success in solving problems is the configuration of the solver, which will be discussed in a separate section in the report. The practical application of the solver will not be ignored either. In this part, we will talk about the specifics of its use in practice, including its comparison with other solvers (in the form allowed for the report).

TOP