Automatic detection of dyslexia based on eye movements during reading in Russian

?

Automatic detection of dyslexia based on eye movements during reading in Russian

P. 59–66.

Laurinavichyute A., Lopukhina A., Reich D.

Dyslexia, a common learning disability, requires an early diagnosis. However, current screening tests are very time- and resourceconsuming. We present an LSTM that aims to automatically classify dyslexia based on eye movements recorded during natural reading combined with basic demographic information and linguistic features. The proposed model reaches an AUC of 0.93 and outperforms the state-of-the-art model by 7 %. We report several ablation studies demonstrating that the fixation features matter the most for classification.

Language: English

Full text

Text on another site

Publication based on the results of:

Neural foundations of language and its impairments: From basic research to clinical practice (2025)

In book

Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics

Vol. 2: Short papers. , Wien: Association for Computational Linguistics, 2025.

Сравнение методов автоматической разметки речевых формул в русскоязычном интернет-дискурсе: пилотное исследование

Попова Т. И., Масленикова А. С., В кн.: Компьютерная лингвистика и интеллектуальные технологии: По материалам ежегодной международной конференции «Диалог». Выпуск 24.Issue 24.: M.: Max press, 2026. С. 420–428.

This study focuses on developing and comparing methods for automatic annotation of speech formulas in a corpus of Russian internet comments. Speech formulas are a class of multiword expressions that convey emotional reactions in dialogue. The research material consisted of a corpus of 10,000 comments (157,261 tokens) collected from five Telegram channels. Dictionary-based formal search ...

Added: June 29, 2026

К ранжированию значимости факторов дестабилизации в странах Азии и Африки методами машинного обучения

Korotayev A., Chernomorchenko I., Медведев И. А., Восток. Афро-азиатские общества: история и современность 2026 № 3 С. 117–130

This study employs machine learning methods to rank factors contributing to large-scale armed and unarmed destabilization across Asian and African countries. Analysis reveals that African nations demonstrate greater vulnerability to armed destabilization (up to full-scale civil wars), whereas Asian countries are more prone to less violent unarmed forms (mass antigovernment demonstrations, riots, general strikes and ...

Added: June 21, 2026

Методология современных лонгитюдных исследований чтения

Bakay E., Antipkina I., Kuznetsova M. I., Отечественная и зарубежная педагогика 2026 Т. 1 № 3(115) С. 6–20

The article analyzes the challenges and prospects of longitudinal research in the field of reading. It examines the systemic limitations of classical longitudinal designs: the problem of unobserved variables, causality, and violations of metric invariance, which lead to erroneous or incomplete interpretation of results. The transition to hybrid research designs is substantiated as a promising ...

Added: June 11, 2026

Artificial intelligence and digital twins for failure prediction in data center cooling systems: a comprehensive literature review (2018–2026)

Butorova A., Bobakov V., Sergeev A. et al., European Physical Journal: Special Topics 2026 P. 1–19

This paper presents a review of artificial intelligence (AI) methods for failure prediction in data center cooling systems, with a focus on the integration of digital twins (DTs), physics-informed learning, and graph-based models. Positioned within complex network science, this review addresses a limitation of conventional graph approaches—their reliance on pairwise connectivity—whereas real-world failures often arise ...

Added: June 10, 2026

Влияние шизофрении на лексический уровень языка

Untila K., Tasenko O., В кн.: Современная лингвистика: ключ к диалогу. Труды и материалы IV Казанского международного лингвистического саммита.Т. 1: СОВРЕМЕННАЯ ЛИНГВИСТИКА: КЛЮЧ К ДИАЛОГУ.: Каз.: Издательство Казанского университета, 2024. С. 221–224.

Шизофрения – это хроническое психическое расстройство, которое выражается как комбинация психотических симптомов – таких как галлюцинации, бред и дезорганизация когнитивных функций. У многих пациентов с диагнозом шизофрения обнаруживаются нарушения речи. Для исследования были отобраны рассказы об истории из жизни из корпуса 3D. В качестве личных историй были собраны ответы на вопросы «Какой самый лучший или запоминающийся ...

Added: June 8, 2026

Proceedings of the 43rd International Conference on Machine Learning (ICML 2026)

Seul: PMLR, 2026.

Added: June 4, 2026

Towards the Ranking of the Importance of Revolutionary Destabilization Factors in Asian and African Countries Using Machine Learning Methods

Chernomorchenko I., Ilya Medvedev, Korotayev A., Cross-Cultural Research 2026 P. 1–49

This study investigates which structural factors most strongly predict armed and unarmed revolutionary destabilization across Sub-Saharan Africa (SSA), the Middle East and North Africa (MENA), as well as Asia using country–year data for 1950–2022 and a set of economic, demographic, political, and climatic indicators. It employs an interpretable machine learning framework (CatBoost with SHAP values ...

Added: June 1, 2026

От неизвестности к прозрачности: обзор технологий объяснимого ИИ (XAI)

Avdoshin S. M., Pesotskaya E. Y., Информационные технологии 2026 Т. 32 № 4 С. 185–194

With the rapid advancement of artificial intelligence, and deep learning in particular, models have emerged that are capable of delivering highly accurate predictions. However, the internal logic of such models remains difficult to interpret—an issue of critical importance, especially in domains where the correctness of an algorithm directly affects high-stakes decision-making. One promising avenue for ...

Added: May 8, 2026

Explainable AI for Industry 5.0: Shedding light on the black box

Avdoshin S. M., Pesotskaya E. Y., Business Informatics 2026 Vol. 20 No. 1 P. 7–28

The rapid development of artificial intelligence (AI) is accompanied by increasing computational complexity and decreasing model transparency, which significantly limits its adoption in critical domains that require a high level of trust, interpretability, and justification of decisions. Under these conditions, the field of Explainable Artificial Intelligence (XAI) has gained particular importance as it focuses on approaches and technologies that ...

Added: May 8, 2026

Современные методы анализа временных рядов в мониторинге и прогнозировании состояния оборудования для механизированной добычи

Neznanov A., Glushko A., Овчинников С. et al., В кн.: Интеллектуальный анализ данных в нефтегазовой отрасли.: М.: ООО «Геомодель Развитие», 2024. С. 140–143.

With the development of monitoring systems, now we have the opportunity to collect key performance indicators of devices in the process of artificial lift. Every day a huge amount of telemetry is generated by our devices, which can be used to forecast the working mode and health state of the equipment after the process of ...

Added: April 29, 2026

Machine Learning Approach to Anticancer Activity Prediction of Transition-Metal Complexes Based on a Large-Scale Experimental Database

Krasnov L., Malikov D., Kiseleva M. et al., Journal of Medicinal Chemistry 2026 Vol. 69 No. 8 P. 8838–8851

In this work, we developed a straightforward data-driven approach to predict the cytotoxicity of metal complexes based entirely on their (metal + ligands) composition. To this end, we have manually curated MetalCytoToxDB─a comprehensive experimental database comprising 26,500 IC50 values for 7050 metal complexes against 754 cell lines from 1921 articles. Based on these, machine learning ...

Added: April 23, 2026

LSTM-модель потребления тепловой энергии в многоэтажном жилом здании

Ершов И. А., Системная инженерия и инфокоммуникации 2025 № 4 С. 11–14

The heat consumption of residential buildings is a stochastic series. It is necessary for the design of thermal energy regulators the creation of a neural network model. In the paper, the model is carried out based on Long Short-Term Memory (LSTM). The high accuracy of reproducing the series was achieved by training the model on ...

Added: April 22, 2026

Алгоритм анализа новостной информации для принятия экономических решений

Чудинова О. С., Первицкая Л. А., Ramenskaya A., Индустриальная экономика 2026 № 1 С. 65–78

This article is devoted to the development of an algorithm for analyzing news information using machine learning methods implemented in Python libraries. The choice of tools used at each stage of the algorithm is justified by calculating metrics for the quality of the solution to the corresponding machine learning problems. The algorithm’s results are presented ...

Added: April 20, 2026

Modeling cosolvent effects on solubility in supercritical CO2 using data-driven approaches

Makarov D. M., Kalikin N., Gurikov P. et al., Journal of Supercritical Fluids 2026 Vol. 235 Article 106979

Supercritical CO2 (scCO2 ) is an environmentally friendly solvent, but its low polarity limits the solubility of polar compounds. Cosolvents are commonly used to enhance solvation capability, yet comprehensive datadriven studies are scarce. We compiled the largest dataset to date — 4401 experimental solubility records with 22 cosolvents for 93 nonionic solutes, plus 4855 records ...

Added: April 19, 2026

Эффективность применения прогнозов волатильности в активных торговых стратегиях институциональных инвесторов на российском рынке акций

Lysenok N., Фундаментальная и прикладная математика 2026 Т. 26 № 3 С. 33–42

This study examines the impact of realized volatility forecasts on the performance of active trading strategies in the Russian equity market. Using a sample of 17 liquid stocks over the period 2014–2026, a hybrid forecasting model is developed that combines HAR-J with gradient boosting; its superiority over the baseline HAR-J specification is confirmed by the ...

Added: April 17, 2026