Three professors from Skoltech: Ivan Oseledets, (CREI for Computational and Data Intensive Science and Engineering, Scientific Computing group), Victor Lempitsky (CREI for Computational and Data Intensive Science and Engineering, Computer Vision group) and Philipp Khaitovich (Center for Data-Intensive Biomedicine and Biotechnology) won grants from Russian Ministry of Education and Science within Federal Targeted Program for Research and Development in Priority Areas of Development of the Russian Scientific and Technological Complex for 2014-2020.
Prof. Ivan Oseledets got a grant for a research entitled “QTT-based technology for solving multiscale problems.” Prof. Oseledets’s group is going to work on this topic in collaboration with Prof. Christoph Schwab group, Swiss Federal Institute of Technology in Zurich (ETHZ)). QTT (named quantized tensor train) is the unique computational tool for solving high-dimensional problems.
The multiscale problems that they consider are any physical problems with periodic/quasiperiodic structure. Such problems appear in photonics, composite material. Researchers plan to develop a “unified” fast method for solving such kind of problems.
Prof. Victor Lempitsky won a grant with a project “Efficient Object-Centric Detection, Indexing, and Spatiotemporal Analysis of Unstructured Dynamic Environments”. His international partner for the grant is Computer Vision and Geometry Group from ETH Zurich (lead by Prof. Marc Pollefeys). Their proposal is about developing new types of representations for 3D scenes that the robots would use to navigate, interact with the world, and anticipate the changes in the scene. Current robots mostly use quite low-level representations such as “point clouds” or “occupancy maps”, which contain the information about the geometric shape of the scene, but do not readily possess information about the semantics, e.g. data on the exact location of the objects. Prof. Victor Lempitsky group will focus on developing more high-level and compact representations of 3d scenes based on objects, and their colleagues from Zurich will study spatio-temporal analysis of such representations (e.g. predicting which objects are moving, or which objects have appeared or disappeared since the last observation of the same scene).
Prof. Philipp Khaitovich got a grant for a research “Regulation of brain aging: transcriptional and epigenetic maps and computational regulatory models of the aging human brain”. Philip Khaitovich’s group is going to work on this project with Konstantin Severinov’s and Raul Gainetdinov teams from Skoltech and Shanghai Institute of Biological Sciences. The plan is to bring together the top-level expertise in the fields of data analysis (P.Khaitovich’s and Chinese team), molecular biology (K.Severinov’s team) and molecular neuroscience (R.Gainetdinov’s team), that will create a synergy necessary to decipher complex regulatory mechanisms guiding human brain aging. The researches aim to generate comprehensive transcriptome and epigenome maps of aging human neocortex. On the basis of these maps, they will reconstruct regulatory network controlling age-dependent changes and create mathematical model, describing this network behavior. The results will allow predicting the effect of physiological or pharmacological intervention directed at specific regulators or network modules. This provides a powerful tool for further work on developing physiological or pharmacological treatment aimed at prevention and even reversal of age-dependent changes in the human brain.