All issues

In this paper, we consider convex stochastic optimization problems arising in machine learning applications (e. g., risk minimization) and mathematical statistics (e. g., maximum likelihood estimation). There are two main approaches to solve such kinds of problems, namely the Stochastic Approximation approach (online approach) and the Sample Average Approximation approach, also known as the Monte Carlo approach, (offline approach). In the offline approach, the problem is replaced by its empirical counterpart (the empirical risk minimization problem). The natural question is how to define the problem sample size, i. e., how many realizations should be sampled so that the quite accurate solution of the empirical problem be the solution of the original problem with the desired precision. This issue is one of the main issues in modern machine learning and optimization. In the last decade, a lot of significant advances were made in these areas to solve convex stochastic optimization problems on the Euclidean balls (or the whole space). In this work, we are based on these advances and study the case of arbitrary balls in the $p$-norms. We also explore the question of how the parameter $p$ affects the estimates of the required number of terms as a function of empirical risk.

In this paper, both convex and saddle point optimization problems are considered. For strongly convex problems, the existing results on the same sample sizes in both approaches (online and offline) were generalized to arbitrary norms. Moreover, it was shown that the strong convexity condition can be weakened: the obtained results are valid for functions satisfying the quadratic growth condition. In the case when this condition is not met, it is proposed to use the regularization of the original problem in an arbitrary norm. In contradistinction to convex problems, saddle point problems are much less studied. For saddle point problems, the sample size was obtained under the condition of $\gamma$-growth of the objective function. When $\gamma = 1$, this condition is the condition of sharp minimum in convex problems. In this article, it was shown that the sample size in the case of a sharp minimum is almost independent of the desired accuracy of the solution of the original problem.

The equilibrium configurations of charged electrons, confined in the hard disk potential, are analysed by means of the hybrid numerical algorithm. The algorithm is based on the interpolation formulas, that are obtained from the analysis of the equilibrium configurations, provided by the variational principle developed in the circular model. The solution of the nonlinear equations of the circular model yields the formation of the shell structure which is composed of the series of rings. Each ring contains a certain number of particles, which decreases as one moves from the boundary ring to the central one. The number of rings depends on the total number of electrons. The interpolation formulas provide the initial configurations for the molecular dynamics calculations. This approach makes it possible to significantly increase the speed at which an equilibrium configuration is reached for an arbitrarily chosen number of particles compared to the Metropolis annealing simulation algorithm and other algorithms based on global optimization methods.

This paper considers various approaches to averaging the generalized travel costs calculated for different modes of travel in the transportation network. The mode of transportation is understood to mean both the mode of transport, for example, a car or public transport, and movement without the use of transport, for example, on foot. The task of calculating the trip matrices includes the task of calculating the total matrices, in other words, estimating the total demand for movements by all modes, as well as the task of splitting the matrices according to the mode, also called modal splitting. To calculate trip matrices, gravitational, entropy and other models are used, in which the probability of movement between zones is estimated based on a certain measure of the distance of these zones from each other. Usually, the generalized cost of moving along the optimal path between zones is used as a distance measure. However, the generalized cost of movement differs for different modes of movement. When calculating the total trip matrices, it becomes necessary to average the generalized costs by modes of movement. The averaging procedure is subject to the natural requirement of monotonicity in all arguments. This requirement is not met by some commonly used averaging methods, for example, averaging with weights. The problem of modal splitting is solved by applying the methods of discrete choice theory. In particular, within the framework of the theory of discrete choice, correct methods have been developed for averaging the utility of alternatives that are monotonic in all arguments. The authors propose some adaptation of the methods of the theory of discrete choice for application to the calculation of the average cost of movements in the gravitational and entropy models. The transfer of averaging formulas from the context of the modal splitting model to the trip matrix calculation model requires the introduction of new parameters and the derivation of conditions for the possible value of these parameters, which was done in this article. The issues of recalibration of the gravitational function, which is necessary when switching to a new averaging method, if the existing function is calibrated taking into account the use of the weighted average cost, were also considered. The proposed methods were implemented on the example of a small fragment of the transport network. The results of calculations are presented, demonstrating the advantage of the proposed methods.

The development of a high-temperature gas-cooled reactor (HTGR) constituting a part of nuclear power-and-process station and intended for large-scale hydrogen production is now in progress in the Russian Federation. One of the key objectives in development of the high-temperature gas-cooled reactor is the computational justification of the accepted design.

The article gives the procedure for the computational analysis of thermal and physical characteristics of the high-temperature gas-cooled reactor. The procedure is based on the use of the state-of-the-art codes for personal computer (PC).

The objective of thermal and physical analysis of the reactor as a whole and of the core in particular was achieved in three stages. The idea of the first stage is to justify the neutron physical characteristics of the block-type core during burn-up with the use of the MCU-HTR code based on the Monte Carlo method. The second and the third stages are intended to study the coolant flow and the temperature condition of the reactor and the core in 3D with the required degree of detailing using the FlowVision and the ANSYS codes.

For the purpose of carrying out the analytical studies the computational models of the reactor flow path and the fuel assembly column were developed.

As per the results of the computational modeling the design of the support columns and the neutron physical characteristics of the fuel assembly were optimized. This results in the reduction of the total hydraulic resistance of the reactor and decrease of the maximum temperature of the fuel elements.

The dependency of the maximum fuel temperature on the value of the power peaking factors determined by the arrangement of the absorber rods and of the compacts of burnable absorber in the fuel assembly is demonstrated.

The work is devoted to the description of the author’s formal statements of the clustering problem for a given number of clusters, algorithms for their solution, as well as the results of using this toolkit in medicine.

The solution of the formulated problems by exact algorithms of implementations of even relatively low dimensions before proving optimality is impossible in a finite time due to their belonging to the NP class.

In this regard, we have proposed a hybrid algorithm that combines the advantages of precise methods based on clustering in paired distances at the initial stage with the speed of methods for solving simplified problems of splitting by cluster centers at the final stage. In the development of this direction, a sequential hybrid clustering algorithm using random search in the paradigm of swarm intelligence has been developed. The article describes it and presents the results of calculations of applied clustering problems.

To determine the effectiveness of the developed tools for optimal clustering of multidimensional objects according to a variety of heterogeneous indicators, a number of computational experiments were performed using data sets including socio-demographic, clinical anamnestic, electroencephalographic and psychometric data on the cognitive status of patients of the cardiology clinic. An experimental proof of the effectiveness of using local search algorithms in the paradigm of swarm intelligence within the framework of a hybrid algorithm for solving optimal clustering problems has been obtained.

The results of the calculations indicate the actual resolution of the main problem of using the discrete optimization apparatus — limiting the available dimensions of task implementations. We have shown that this problem is eliminated while maintaining an acceptable proximity of the clustering results to the optimal ones. The applied significance of the obtained clustering results is also due to the fact that the developed optimal clustering toolkit is supplemented by an assessment of the stability of the formed clusters, which allows for known factors (the presence of stenosis or older age) to additionally identify those patients whose cognitive resources are insufficient to overcome the influence of surgical anesthesia, as a result of which there is a unidirectional effect of postoperative deterioration of complex visual-motor reaction, attention and memory. This effect indicates the possibility of differentiating the classification of patients using the proposed tools.

In today’s technology-driven world, lite devices like Internet of Things (IoT) devices and microcontrollers (MCUs) are becoming increasingly common. These devices are more energyefficient and affordable, often with reduced features compared to the standard versions such as very limited memory and processing power for typical machine learning models. However, modern machine learning models can have millions of parameters, resulting in a large memory footprint. This complexity not only makes it difficult to deploy these large models on resource constrained devices but also increases the risk of latency and inefficiency in processing, which is crucial in some cases where real-time responses are required such as autonomous driving and medical diagnostics. In recent years, neural networks have seen significant advancements in model optimization techniques that help deployment and inference on these small devices. This narrative review offers a thorough examination of the progression and latest developments in neural network optimization, focusing on key areas such as quantization, pruning, knowledge distillation, and neural architecture search. It examines how these algorithmic solutions have progressed and how new approaches have improved upon the existing techniques making neural networks more efficient. This review is designed for machine learning researchers, practitioners, and engineers who may be unfamiliar with these methods but wish to explore the available techniques. It highlights ongoing research in optimizing networks for achieving better performance, lowering energy consumption, and enabling faster training times, all of which play an important role in the continued scalability of neural networks. Additionally, it identifies gaps in current research and provides a foundation for future studies, aiming to enhance the applicability and effectiveness of existing optimization strategies.

We developed a model of honeybee colony foraging based on reaction – diffusion equations. Employed bees transmit information about their food sources using dance, and job seekers in the hive can choose any dance they like and thus join the exploitation of the corresponding source. We consider two strategies of dance selection: a targeted one, when bees analyze information on the dance floor and choose the most energetic and longest dance corresponding to the most profitable source, and a simple random choice of the first dance they encounter. Modelling showed that the greatest profit (food influx into the hive) is provided by the random choice of dance, as paradoxical as it may seem at first glance. Optimization of profit by each agent for itself (targeted choice of dances) is rather a disadvantage for the colony, and “non-optimality” in dance choice can be the result of useful evolutionary adaptation.

This paper investigates the effectiveness of Retrieval-Augmented Generation (RAG) combined with various Large Language Models (LLMs) for document retrieval and information access in corporate information systems. We survey typical use-cases of LLMs in enterprise environments, outline the RAG architecture, and discuss the major challenges that arise when integrating LLMs into a RAG pipeline. A system architecture is proposed that couples a text-vector encoder with an LLM. The encoder builds a vector database that indexes a library of corporate documents. For every user query, relevant contextual fragments are retrieved from this library via the FAISS engine and appended to the prompt given to the LLM. The LLM then generates an answer grounded in the supplied context. The overall structure and workflow of the proposed RAG solution are described in detail. To justify the choice of the generative component, we benchmark a set of widely used LLMs — ChatGPT, GigaChat, YandexGPT, Llama, Mistral, Qwen, and others — when employed as the answer-generation module. Using an expert-annotated test set of queries, we evaluate the accuracy, completeness, linguistic quality, and conciseness of the responses. Model-specific characteristics and average response latencies are analysed; the study highlights the significant influence of available GPU memory on the throughput of local LLM deployments. An overall ranking of the models is derived from an aggregated quality metric. The results confirm that the proposed RAG architecture provides efficient document retrieval and information delivery in corporate environments. Future research directions include richer context augmentation techniques and a transition toward agent-based LLM architectures. The paper concludes with practical recommendations on selecting an optimal RAG–LLM configuration to ensure fast and precise access to enterprise knowledge assets.

Amid the ongoing trend of rapid urbanization and the intensive development of megacities and large cities worldwide, the impact of man-made vibrations on residential structures and infrastructure is increasing. The operation of subway systems, construction using pile-driving and drilling equipment, and heavy traffic have become active sources of wave disturbances, which can be a decisive factor in reducing the structural stability of buildings and, consequently, their long-term reliability. This paper proposes a numerical calculation using the grid-characteristic method to model elastic waves propagating through soil layers and load-bearing structures from various sources. By solving the direct problem of numerical pulse simulation and varying its location, the values of velocity vector projections and components of the Cauchy stress tensor were obtained at each time step. Two scenarios were examined: the first simulates the impact of noise generated by construction work or nearby traffic, while the second demonstrates how a subway running through an underground tunnel affects multi-story residential buildings. Wave propagation patterns from these sources were visualized in terms of the parameters of interest, enabling a quick and convenient comprehensive analysis of the problem. The analysis of the obtained data will help adjust the timing and types of repair work, identify structural weak points, and develop innovative methods for preserving historical buildings that are cultural heritage sites. Additionally, it will allow for the most economically optimal construction of modern buildings near architectural landmarks, provide an efficient and safe action plan in emergencies, and modernize existing construction technologies to enhance the comfort of residential buildings, office structures, and other socially significant facilities. It will also aid in selecting the most suitable locations for modern high-precision manufacturing plants.

The article is dedicated to mathematic modeling of natural water desalination process by method of thermal distillation. The article gives the equations which allow describing the processes of film flowing and boiling of water, steam condensation and vacuum maintenance. The article presents the algorithm of calculation, implemented in MatLab computer mathematic system and Excel electronic tables, and the initial data required for the calculation. The model has been checked for adequacy. The calculation of ten-effect distillation system is given. The results of work can be used in design and optimization of process conditions for distillation systems.