LibreTimes

Search results for “topoi”

47 results

Incremental Learning of Topic Models for Finding Trend Topics in Scientific Publications

2022Journal articleN. A. Gerasimenko, A. S. Chernyavsky, M. A. Nikiforova, M. D. Nikitin +1

Doklady Mathematics

With a soaring number of scientific publications and rapid emergence of new directions and approaches, the scientific community faces the task of timely identification of trends. By a trend, we mean a semantically homogeneous topic characterized by a steady lexical kernel and a sharp, often exponential increase in the number of publications [1]. Examples of trends in machine learning are “LSTM,” “deep learning,” “word2vec,” “BERT,” and “fake news detection.” For real-time detection of trend topics from a stream of scientific publications, we use incremental methods of probabilistic topic modeling. An ARTM-based approach to early trend detection has been shown to outperform popular classical and neural network approaches to this task. A dataset of 91 trends for performance evaluation has been manually collected and made available for public use.

0
1

Convergence of the Algorithm of Additive Regularization of Topic Models

2021Journal articleI. A. Irkhin, K. V. Vorontsov

Proceedings of the Steklov Institute of Mathematics

The problem of probabilistic topic modeling is as follows. Given a collection of text documents, find the conditional distribution over topics for each document and the conditional distribution over words (or terms) for each topic. Log-likelihood maximization is used to solve this problem. The problem generally has an infinite set of solutions and is ill-posed according to Hadamard. In the framework of Additive Regularization of Topic Models (ARTM), a weighted sum of regularization criteria is added to the main log-likelihood criterion. The numerical method for solving this optimization problem is a kind of an iterative EM-algorithm written in a general form for an arbitrary smooth regularizer as well as for a linear combination of smooth regularizers. This paper studies the problem of convergence of the EM iterative process. Sufficient conditions are obtained for the convergence to a stationary point of the regularized log-likelihood. The constraints imposed on the regularizer are not too restrictive. We give their interpretations from the point of view of the practical implementation of the algorithm. A modification of the algorithm is proposed that improves the convergence without additional time and memory costs. Experiments on a news text collection have shown that our modification both accelerates the convergence and improves the value of the criterion to be optimized.
0
1

Interpretable probabilistic embeddings: bridging the gap between topic\n models and neural networks

2017PreprintPotapenko, Anna, Popov, Artem, Vorontsov, Konstantin

arXiv (Cornell University)

We consider probabilistic topic models and more recent word embedding\ntechniques from a perspective of learning hidden semantic representations.\nInspired by a striking similarity of the two approaches, we merge them and\nlearn probabilistic embeddings with online EM-algorithm on word co-occurrence\ndata. The resulting embeddings perform on par with Skip-Gram Negative Sampling\n(SGNS) on word similarity tasks and benefit in the interpretability of the\ncomponents. Next, we learn probabilistic document embeddings that outperform\nparagraph2vec on a document similarity task and require less memory and time\nfor training. Finally, we employ multimodal Additive Regularization of Topic\nModels (ARTM) to obtain a high sparsity and learn embeddings for other\nmodalities, such as timestamps and categories. We observe further improvement\nof word similarity performance and meaningful inter-modality similarities.\n

0
1

Optimizing Modality Weights in Topic Models of Transactional Data

2022Journal articleK. Ya. Khrylchenko, K. V. Vorontsov

Automation and Remote Control

Modern natural language processing models such as transformers operate multimodal data. In the present paper, multimodal data is explored using multimodal topic modeling on transactional data of bank corporate clients. A definition of the importance of modality for the model is proposed on the basis of which improvements are considered for two modeling scenarios: preserving the maximum amount of information by balancing modalities and automatic selection of modality weights to optimize auxiliary criteria based on topic representations of documents. A model is proposed for adding numerical data to topic models in the form of modalities: each topic is assigned a normal distribution with learning parameters. Significant improvements are demonstrated in comparison with standard topic models on the problem of modeling bank corporate clients. Based on the topic representations of the bank’s customers, a 90-day delay on the loan is predicted.
0
1

Mining Ethnic Content Online with Additively Regularized Topic Models

2016Journal articleMurat Apishev, Sergei Koltcov, Olessia Koltsova, Sergey Nikolenko +1

Computación y Sistemas

Social studies of the Internet have adopted large-scale text mining for unsupervised discovery of topics related to specific subjects. A recently developed approach to topic modeling, additive regularization of topic models (ARTM), provides fast inference and more control over the topics with a wide variety of possible regularizers than developing LDA extensions. We apply ARTM to mining ethnic-related content from Russian-language blogosphere, introduce a new combined regularizer, and compare models derived from ARTM with LDA. We show with human evaluations that ARTM is better for mining topics on specific subjects, finding more relevant topics of higher or comparable quality.
0
1

Regularization, robustness and sparsity of probabilistic topic models

2012Journal articleKonstantin Vyacheslavovich Vorontsov, Anna Alexandrovna Potapenko

Computer Research and Modeling

We propose a generalized probabilistic topic model of text corpora which can incorporate heuristics of Bayesian regularization, sampling, frequent parameters update, and robustness in any combinations. Wellknown models PLSA, LDA, CVB0, SWB, and many others can be considered as special cases of the proposed broad family of models. We propose the robust PLSA model and show that it is more sparse and performs better that regularized models like LDA.
0
1

Топология-2. Лекция 1

2026LectureФёдор Вылегжанин

Зачем нужны гомологии — фундаментальная группа как функтор, гомотопические группы, бордизмы и их комбинаторная аппроксимация; цепные комплексы и их гомологии; абстрактные симплициальные комплексы, геометрическая реализация и триангуляции; комплекс симплициальных цепей, доказательство d2=0d^2 = 0, симплициальные гомологии отрезка, двух точек и границы треугольника; приведённые гомологии.

0
00

Топология-2. Лекция 2

2026LectureФёдор Вылегжанин

Где симплициальные гомологии встречаются в жизни — топологический анализ данных. Объединение шаров вокруг облака точек, персистентные модули и их разложение в сумму интервальных модулей над полем (структурная теорема и алгоритм), баркод; комплексы Виеториса–Рипса и Чеха; теорема об устойчивости баркода, расстояния Хаусдорфа, Громова–Хаусдорфа и Васерштейна; открытые покрытия и их нерв, покрытия Лере и теорема о нерве, категория покрытий, паракомпактность, размерность по Лебегу, гомологии и когомологии Чеха.
0
00

Топология-2. Лекция 3

2026LectureФёдор Вылегжанин

Сингулярные гомологии — стандартные симплексы и вложения гиперграней, сингулярные симплексы и цепи, дифференциал; нулевые гомологии и компоненты линейной связности, гомологии точки, приведённые гомологии; цепные отображения и отображения гомологий, функториальность, цепное отображение, индуцированное непрерывным отображением; цепные гомотопии и призма; теорема о гомотопической инвариантности с полным доказательством через призменный оператор; подкомплексы, факторкомплексы и гомологии пары.
0
00

Топология-2. Листок 1

2026Problem sheetФёдор Вылегжанин

Задачи курса «Топология-2» (Ф. Е. Вылегжанин, НМУ, осень 2026), листок 1 от 9 сентября 2026 г. — симплициальные гомологии и цепные комплексы. Нулевые гомологии и компоненты связности; гомологии несвязного объединения и букета; триангуляции окружности, сферы, ленты Мёбиуса; джойн и произведение симплициальных комплексов; эйлерова характеристика и неравенства Морса; число вершин триангуляции поверхности; операции над цепными комплексами; точные последовательности; нормальная форма Смита и разложение комплекса.
0
00

Топология-2. Листок 2

2026Problem sheetФёдор Вылегжанин

Задачи курса «Топология-2» (Ф. Е. Вылегжанин, НМУ, осень 2026), листок 2 от 16 сентября 2026 г. — сингулярные гомологии и цепные гомотопии. Цепное отображение из симплициальных цепей в сингулярные; ретракция и прямое слагаемое; гомоморфизм Гуревича из фундаментальной группы в первые гомологии; конечные CW-комплексы и классы гомологий; гомологии объединения цепочки пространств; гомологии симплекса и его остова; симплициальные отображения; цепная гомотопическая эквивалентность комплекса и его гомологий; цепная гомотопность как отношение эквивалентности; 5-лемма.
0
00

Towards free-space ultrastable optical frequency transfer

2023Conference talkКсения Лискова, Alexey Legoshin, K. S. Kudeyarov, G. A. Vishnyakova +5

VII International Conference on Quantum Technologies (ICQT 2023)

Nowadays optical frequency transfer has become an essential component for numerous quantum technology applications. There is a gradual increase in demand for such systems due to the rapid development of quantum technologies themselves, e. g. optical frequency transmission became crucial for high-precision timekeeping and communication systems relying on the frequency stability of quantum clocks that offer unparalleled accuracy, making ideal for metrology, navigation, and testing fundamental physical theories. Moreover, optical frequency transmission has already found popularity in such up-to-date topics as quantum communication and quantum cryptography.

For over 6 years, our laboratory has been developing various systems for the transmission of stabilized frequency signals. In 2017, we started with the elaboration of a phase noise compensation system in a 5-meter long optical fiber frequency transmission line and afterwards extended the line to 2.8 kilometers. Taking advantage of the result we also connected three of our laboratories with fiber optic cables with phase noise compensation systems to facilitate the process of comparing developed frequency standards. With an awareness of the fiber optic links applicability limitations (such as insufficient mobility and flexibility) we developed a 5-meter long free-space optical link with the same phase-noise compensation system in 2020 and increased its length to 17 meters with the addition of a precision pointing system in 2021.

Finally, in 2023, we introduced a 215-cm free-space optical transmission link with a flexible pointing system that enables dynamically stable tracking of moving objects: potentially drones or even satellites. The test scheme of this transfer system with defined upgrades is presented in Figure 1.

Phase noise compensation system for both fiber and free-space optical links The transmission link introduces phase noise into the signal. To compensate for corresponding frequency shifts, a laser beam used for transmission is split into two parts. The first part passes through a reference arm of an interferometer, while the second is transmitted to the receiver via an acousto-optic modulator (AOM1) and partly reflected back. The returned signal contains doubled link noise and is heterodyned with the reference beam. The resulting beat signal is used in a phase-locked loop that controls the frequency shift introduced by AOM1 and compensates the link noise.

Pointing system for free-space optical link To compensate for small high-frequency beam direction fluctuations, a precise (fast) pointing system was used, which includes a position-sensitive quadrant photodetector managing a mirror with two-coordinate galvanic control. To expand the pointing range, we implement a coarse (slow) pointing system by attaching the optical plate to an alt-azimuth telescope mount. The servo signal of the fast system is used to correct the angular velocity of the dynamic tripod rotation around two axes.

In the near future, we plan to extend the free-space link length and test the system in actual environments when pointing at a moving UAV. In addition, we plan to focus on processing the received noise signal in order to use it to obtain information about atmospheric parameters.

We strongly believe that the development of optical transmission systems is extremely useful for the expansion of quantum technologies.

0
2

Reinforcement Networks: novel framework for collaborative Multi-Agent Reinforcement Learning tasks

2025PreprintKryzhanovskiy, Maksim, Glazyrina, Svetlana, Ischenko, Roman, Vorontsov, Konstantin

arXiv (Cornell University)

Modern AI systems often comprise multiple learnable components that can be naturally organized as graphs. A central challenge is the end-to-end training of such systems without restrictive architectural or training assumptions. Such tasks fit the theory and approaches of the collaborative Multi-Agent Reinforcement Learning (MARL) field. We introduce Reinforcement Networks, a general framework for MARL that organizes agents as vertices in a directed acyclic graph (DAG). This structure extends hierarchical RL to arbitrary DAGs, enabling flexible credit assignment and scalable coordination while avoiding strict topologies, fully centralized training, and other limitations of current approaches. We formalize training and inference methods for the Reinforcement Networks framework and connect it to the LevelEnv concept to support reproducible construction, training, and evaluation. We demonstrate the effectiveness of our approach on several collaborative MARL setups by developing several Reinforcement Networks models that achieve improved performance over standard MARL baselines. Beyond empirical gains, Reinforcement Networks unify hierarchical, modular, and graph-structured views of MARL, opening a principled path toward designing and training complex multi-agent systems. We conclude with theoretical and practical directions - richer graph morphologies, compositional curricula, and graph-aware exploration. That positions Reinforcement Networks as a foundation for a new line of research in scalable, structured MARL.
0
1