?
Automatic differentiation for Riemannian optimization on low-rank matrix and tensor-train manifolds
SIAM Journal of Scientific Computing. 2022. Vol. 44. No. 2. P. A843–A869.
In scientific computing and machine learning applications, matrices and more general multidimensional arrays (tensors) can often be approximated with the help of low-rank decompositions. Since matrices and tensors of fixed rank form smooth Riemannian manifolds, one of the popular tools for finding low-rank approximations is to use Riemannian optimization. Nevertheless, efficient implementation of Riemannian gradients and Hessians, required in Riemannian optimization algorithms, can be a nontrivial task in practice. Moreover, in some cases, analytic formulas are not even available. In this paper, we build upon automatic differentiation and propose a method that, given an implementation of the function to be minimized, efficiently computes Riemannian gradients and matrix-by-vector products between an approximate Riemannian Hessian and a given vector.
Nozdrinova E., Pochinka O., Shmukler V., Математический сборник 2026 Т. 217 № 6 С. 71–89
Гомеоморфизмы топологических пространств называются эквивалентными по надстройке, если надстройки над ними топологически эквивалентны. В частности, топологически сопряженные гомеоморфизмы эквивалентны по надстройке. Известно, что для гомологически неприводимых гомеоморфизмов их топологическая сопряженность является необходимым и достаточным условием их эквивалентности по надстройке. Тогда как инварианты топологической сопряженности гомологически приводимых гомеоморфизмов во многих случаях являются избыточными для эквивалентности по ...
Added: June 3, 2026
Gnetov F., Konakov V., Успехи математических наук 2026 Т. 81 № 3 (489) С. 161–162
Пусть M обозначает симметрическое пространство некомпактного типа ранга 1. Опираясь на фундаментальную работу [1], в [2] было показано, что плотность соответствующим образом нормированной суммы независимых Hn-значных случайных величин, определенная через сложение Мёбиуса в модели шара Пуанкаре, сходится к фундаментальному решению соответствующего уравнения теплопроводности. Пределом являлся нормальный закон на Hn, соответствующий ядру теплопроводности, определяемому оператором Лапласа–Бельтрами. ...
Added: June 2, 2026
Silakov D., Системный администратор 2026 № 3 С. 28–33
В статье про платформы для разработки открытого ПО в Китае мы рассказали про GitCode – молодой проект, позиционируемый как площадка для разработчиков со всего мира. Сейчас на GitCode размещаются проекты, созданные в КНР, но некоторые из них уже известны и на международной арене. Помочь открытым проектам в становлении, развитии и расширению аудитории призван фонд OpenAtom ...
Added: June 2, 2026
Slivnitsin P., Mylnikov L., Engineering Applications of Artificial Intelligence 2026 Vol. 179 Article 115185
The paper describes a applied artificial intelligence task of recognition-by-components method of real objects based on the recognition of a limited set of primitives or components. The recognition-by-components makes it possible to determine the components, that compose an object, and increase the number of recognizable objects without degrading the recognition quality. Training is performed on ...
Added: May 29, 2026
Gorbounov Vassily, Kazakov A., Data Analytics and Topology 2025 Vol. 1 No. 1 P. 33–45
A classic problem in data analysis is studying the systems of subsets defined by either a similarity or a dissimilarity function on X which is either observed directly or derived from a data set.
For an electrical network there are two functions on the set of the nodes defined by the resistance matrix and the response ...
Added: May 28, 2026
Mokienko O., Zisman M. A., Bobrov P. et al., American Journal of Physical Medicine and Rehabilitation 2026 Vol. 105 No. 6 P. 555–563
Brain-computer interfaces (BCIs) represent a promising technology for restoring lower limb motor functions and gait after stroke. The application of BCIs in this field is supported by a limited number of studies. The objective of the review was to systematically and critically evaluate the current evidence on the use of BCIs for lower limb function ...
Added: May 28, 2026
М.: Институт проблем управления им. В.А. Трапезникова РАН, 2024.
В сборник вошли материалы VIII Международной научной конференции «Информационные технологии и технические средства управления» (ICCT-2024). На конференции были рассмотрены вопросы, касающиеся перспектив развития научного приборостроения в телекоммуникационных и управляющих системах, биомедицинской информатики, аппаратного и программного обеспечения информационнокоммуникационных систем, надежности, диагностики и неразрушающего контроля, систем управления и автоматизации, цифровых экосистем, управления производством и логистикой, методов математического ...
Added: May 27, 2026
Degtyarev A., Bakhurin S., Yudin N., DSPA 2026 P. 1–6
This paper investigates one possible solution to the problem of self-interference cancellation (SIC) arising in the design of in-band full-duplex (IBFD) communication systems. Self-interference cancellation is performed in the digital domain using multilayer nonlinear models adapted via gradient-based optimization. The presence of local minima and saddle points during the adaptation of multilayer models limits the ...
Added: May 26, 2026
Ilyashenko Y., Shilin I., Stanislav Minkov, Russian Journal of Mathematical Physics 2026 Vol. 33 No. 1 P. 89–106
In this paper, new numerical invariants of structurally unstable vector fields in the plane
are found. One of the main tools is an improved asymptotics of sparkling saddle connections that
occur when a separatrix loop of a hyperbolic saddle breaks. Another main tool is a new topological
invariant of two arithmetic progressions, both perturbed and unperturbed, on the ...
Added: May 26, 2026
Androsov I., Proceedings of the Institute for System Programming of the RAS 2026 Vol. 38 No. 3 P. 87–114
This paper examines echo state networks (ESNs), one of the most prevalent approaches to
implementing reservoir computing. An ESN consists of a recurrent neural network with fixed (untrained)
weights and a readout layer that is typically linear and trainable. This approach enables the creation of energyefficient and computationally efficient neural networks capable of real-time learning. However, since ...
Added: May 26, 2026
Gusev I., Maksaev A., Promyslov V., Journal of Mathematical Sciences 2025 Vol. 299 No. 6
The regular graph of the space of n × m matrices over a field F is defined as the undirected graph whose vertices are matrices of rank min(n, m), and distinct matrices A and B are connected by an edge if and only if rk(A + B) < min(n, m). In this paper, for |F| ...
Added: May 25, 2026
Караваева Е. А., Кулигин Л. А., Rezunik L. et al., Труды Института системного программирования РАН 2026 Т. 38 № 3 С. 67–94
В статье представлен метод рефакторинга исходного кода на основе интеграции большой языковой модели (LLM) и расширенной UML-модели программного кода. Предложенный подход позволяет выявлять проблемные участки кода с использованием функций тревожности и структурных метрик классов, а затем выполнять автоматизированный рефакторинг. Ключевой особенностью метода является использование LLM для генерации формальных спецификаций на языке OCL (Object Constraint Language), ...
Added: May 24, 2026
Tyukin I., Tyukina T., van Helden D. P. et al., Information Sciences 2024 Vol. 678 Article 120856
AI errors pose a significant challenge, hindering real-world applications. This work introduces a novel approach to cope with AI errors using weakly supervised error correctors that guarantee a specific level of error reduction. Our correctors have low computational cost and can be used to decide whether to abstain from making an unsafe classification. We provide ...
Added: May 23, 2026
Zaikin A., Sviridov I., Sosedka A. et al., Technologies 2026 Vol. 14 No. 2 Article 84
High-dimensional tabular data are common in biomedical and clinical research, yet conventional machine learning methods often struggle in such settings due to data scarcity, feature redundancy, and limited generalization. In this study, we systematically evaluate Synolitic Graph Neural Networks (SGNNs), a framework that transforms high-dimensional samples into sample-specific graphs by training ensembles of low-dimensional pairwise ...
Added: May 23, 2026
Alexander Molozhavenko, Rakhuba M., Computational and Applied Mathematics 2026 Vol. 45 No. 6 Article 221
This paper studies tensors that admit decomposition in the Extended Tensor Train (ETT) format, with a key focus on the case where some decomposition factors are constrained to be equal. This factor sharing introduces additional challenges, as it breaks the multilinear structure of the decomposition. Nevertheless, we show that Riemannian optimization methods can naturally handle ...
Added: December 22, 2025
Pospelov N., Chertkov A., Beketov M. et al., Neurocomputing 2025 Vol. 618 Article 129070
Elements of neural networks, both biological and artificial, can be described by their selectivity for specific cognitive features. Understanding these features is important for understanding the inner workings of neural networks. For a living system, such as a neuron, whose response to a stimulus is unknown and not differentiable, the only way to reveal these ...
Added: December 14, 2024
Peshekhonov I., Aleksey Arzhantsev, Rakhuba M., , in: Proceedings of The 27th International Conference on Artificial Intelligence and Statistics (AISTATS 2024), 2-4 May 2024, Palau de Congressos, Valencia, Spain. PMLR: Volume 238Vol. 238.: Valencia: PMLR, 2024. Ch. 238 P. 3304–3312.
Added: November 29, 2024
Shalileh S., , in: Data Analysis and Optimization. In Honor of Boris Mirkin's 80th Birthday.: Springer, 2023. P. 349–361.
Based upon variational inference (VI) a new set of classification algorithms has recently emerged. This set of algorithms aims (A) to increase generalization power, (B) to decrease computational complexity. However, the complex math and implementation considerations have led to the emergence of black-box variational inference methods (BBVI). Relying on these principles, we assume the existence ...
Added: October 17, 2023
Marcati C., Rakhuba M., Ulander J. ., Calcolo 2022 Article 2
We derive rank bounds on the quantized tensor train (QTT) compressed approximation of singularly perturbed reaction diffusion boundary value problems in one dimension. Specifically, we show that, independently of the scale of the singular perturbation parameter, a numerical solution with accuracy 0 < 𝜀 < 1 can be represented in the QTT format with a number of parameters that ...
Added: October 31, 2021
Obukhov A., Rakhuba M., Liniger A. et al., , in: International Conference on Artificial Intelligence and Statistics, 13-15 April 2021, VirtualVol. 130.: PMLR, 2021. P. 3547–3555.
Added: October 31, 2021
Высоцкий Л. И., Журнал вычислительной математики и математической физики 2021 Т. 61 № 5 С. 776–786
Tensorizations of functions are studied, that is, tensors with elements 𝐴(𝑖1,…,𝑖𝑑)=𝑓(𝑥(𝑖1,…,𝑖𝑑)) 𝐴(𝑖1,…,𝑖𝑑)=𝑓(𝑥(𝑖1,…,𝑖𝑑)), where 𝑓(𝑥) is some function defined on an interval and {𝑥(𝑖1,…,𝑖𝑑)} is a grid on this interval. For tensors of this type, the problem of approximation by tensors admitting a tensor train (ТТ) decomposition with low ТТ ranks is posed. For the class ...
Added: October 31, 2021
Высоцкий Л. И., Смирнов А. В., Тыртышников Е. Е., Lobachevskii Journal of Mathematics 2021 Vol. 42 P. 1608–1621
Numerical integration is a classical problem emerging in many fields of science. Multivariate integration cannot be approached with classical methods due to the exponential growth of the number of quadrature nodes. We propose a method to overcome this problem. Tensor-train decomposition of a tensor approximating the integrand is constructed and used to evaluate a multivariate ...
Added: October 31, 2021
IEEE, 2020.
Dimensionality reduction problem is stated as finding a mapping f:X ∈ R m → Z ∈ R n , where ≪ m while preserving some relevant properties of the data. We formulate topology-preserving dimensionality reduction as finding the optimal orthogonal projection to the lower-dimensional subspace which minimizes discrepancy between persistent diagrams of the original data and the projection. This ...
Added: October 14, 2021