“统数论坛”2026系列报告17：俄罗斯科学院院士Evgeny Tyrtyshnikov学术报告

Tensor decompositions become a very popular tool for modelling data in many application problems. We discuss some still open issues about the rank-bounded sets for the canonical polyadic decomposition and new developments of cross-approximation approach to optimization problems with the tensor-train model. Evgeny Tyrtyshnikov is an academician of the Russian Academy of Sciences (elected in 2016), the Chairman of the National Committee for Industrial and Applied Mathematics. He is a leading researcher at the Marchuk Institute of Numerical Mathematics of the Russian Academy of Sciences. Professor and Chairman at the Lomonosov Moscow State University (since 2004). Best teacher of the year award at Lomonosov Moscow State University (2014). Laureate of the Sber Scientific Prize in the nomination “Digital Universe” (2023). Yangtze Professor at the Shenzhen-Moscow-Beijing University (since November 2024). Lingnan Fellow in the Lingnan University Institute for Advanced Study at Hong Kong. Author of 8 books and more than 120 papers. Several collaborations with industry (Cray Research, Baker Hughes, Morgan Stanley, Huawei and others). Editor-in-chief of the Journal of Computational Mathematics and Mathematical Physics. Member of editorial boards of several journals: Calcolo, Sbornik Mathematics, Journal of Numerical Mathematics, Russian Journal of Numerical Analysis and Mathematical Modelling, Siberian Journal of Numerical Mathematics, Lobachevsky Journal of Mathematics. His research interests include multi-dimensional problems, structured matrices, asymptotic matrix analysis, spectral distributions, tensor decompositions and nonlinear approximations in linear algebra and numerical analysis.