![]() |
ИСТИНА |
Войти в систему Регистрация |
ИПМех РАН |
||
The target groups of the school are students and young researchers interested in modern trends in data analytics and in approximation, including big data processing, new machine learning, data mining techniques and rational approximation with orthogonal polynomials. The proposed workshop is devoted to the development of mathematical aspects of data mining methods related with approximation and network (graph based) models. Graph and network models are our days ubiquitous in different fields of science and became more and more popular in approximations, data mining and machine learning. However, it is recognized that there is a lack of mathematical foundations for many approaches and methods of data analytics. Different aspects of network models in approximations, data mining and machine learning will be discussed. The topics of the workshop are rational approximations defined by multi-indices on an integer lattice, multi-level interpolations, reproducing kernels and operator theory on graphs. In addition, it is planned to discuss optimization of computational graph for deep learning, robustness of network models algorithms in machine learning and data mining, network models under uncertainty, multi-level network models and others. Lecturers at the school are internationally recognized experts in the field. Young researchers will have an opportunity to learn and understand a modern theoretical approaches and practical techniques in data analysis and approximation.