Russian Algorithm Ranks Among the World’s Top Machine Learning Tools
Thanks to its high accuracy and automated data handling, the algorithm has become an industry standard, used across Russian business and scientific research.
A Russian-developed machine learning tool, Yandex’s CatBoost, has been ranked among the world’s five most popular non-US machine learning tools. The ranking comes from the ML Global Impact Report 2025 published by Marktechpost and cited by Forbes.
The algorithm rivals tools from major IT players such as Google, Intel, and Microsoft in terms of how often it is cited in scientific papers. It is used by researchers in 51 countries, including the United States, China, Saudi Arabia, and other global leaders in science.
CatBoost’s key strength is its efficient handling of tabular data for real-world tasks such as search ranking, weather forecasting, and medical diagnostics. The algorithm helps analyze structured datasets for decision-making, uncovering complex and non-obvious patterns hidden in tables.
Serving Business and Science
The tool has been openly available since 2017 and is widely used not only in Yandex services but also across major Russian companies and in academic research.
For example, CatBoost has been integrated into Avito’s multi-level machine learning architecture. In real time, it ranks listings in recommendation systems, predicts conversions within promotion models, and supports content personalization and optimization mechanisms.
Researchers at the Moscow Institute of Physics and Technology have used CatBoost to develop methods for predicting the condition of lithium-ion batteries. The algorithm has also been applied to training experimental digital twins of batteries on synthetic data, enabling forecasts of various battery properties.
