LibreTimes

Search results for “topoi”

44 results

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
3

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
5

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
3

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
3

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
5

Topology 2. Lecture 1

2026LectureФёдор Вылегжанин

Why homology is needed — the fundamental group as a functor, homotopy groups, bordism and its combinatorial approximation; chain complexes and their homology; abstract simplicial complexes, geometric realisation and triangulations; the complex of simplicial chains, the proof of d2=0d^2 = 0, the simplicial homology of a segment, of two points and of the boundary of a triangle; reduced homology.

1
470

Topology 2. Lecture 2

2026LectureФёдор Вылегжанин

Where simplicial homology shows up in real life — topological data analysis. The union of balls around a point cloud, persistence modules and their decomposition into a sum of interval modules over a field (the structure theorem and the algorithm), the barcode; the Vietoris–Rips and Čech complexes; the stability theorem for barcodes, the Hausdorff, Gromov–Hausdorff and Wasserstein distances; open covers and their nerves, Leray covers and the nerve theorem, the category of covers, paracompactness, Lebesgue covering dimension, Čech homology and cohomology.
1
520

Topology 2. Lecture 3

2026LectureФёдор Вылегжанин

Singular homology — standard simplices and the inclusions of facets, singular simplices and chains, the differential; zeroth homology and path components, the homology of a point, reduced homology; chain maps and maps of homology, functoriality, the chain map induced by a continuous map; chain homotopies and the prism; the homotopy invariance theorem with a full proof via the prism operator; subcomplexes, quotient complexes and the homology of a pair.
1
550

Topology 2. Problem sheet 1

2026Problem sheetФёдор Вылегжанин

Problems for the course "Topology 2" (F. E. Vylegzhanin, Independent University of Moscow, autumn 2026), sheet 1 of 9 September 2026 — simplicial homology and chain complexes. Zeroth homology and connected components; homology of a disjoint union and of a wedge; triangulations of the circle, the sphere and the Möbius band; the join and the product of simplicial complexes; the Euler characteristic and the Morse inequalities; the number of vertices of a triangulated surface; operations on chain complexes; exact sequences; the Smith normal form and the splitting of a complex.
1
450

Topology 2. Problem sheet 2

2026Problem sheetФёдор Вылегжанин

Problems for the course "Topology 2" (F. E. Vylegzhanin, Independent University of Moscow, autumn 2026), sheet 2 of 16 September 2026 — singular homology and chain homotopies. A chain map from simplicial to singular chains; retractions and direct summands; the Hurewicz homomorphism from the fundamental group to the first homology; finite CW complexes and homology classes; homology of a union of a chain of spaces; homology of a simplex and of its skeleton; simplicial maps; the chain homotopy equivalence between a complex and its homology; chain homotopy as an equivalence relation; the five lemma.
1
400

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
4

An example of a non-reduced Picard scheme

2026TheorySean Cotner

An example, due to Serre, of a smooth projective surface in positive characteristic whose Picard scheme fails to be reduced, worked out via the two governing dimension inequalities.
1
170

Тематическое моделирование

2026LectureКонстантин Вячеславович Воронцов

Тематическое моделирование. Лекция курса «Машинное обучение» К. В. Воронцова. Задача тематического моделирования коллекции текстовых документов. Метод максимума правдоподобия.
0
110