{"status":"ok","elements":"
ADASE<\/strong> group research is about developing efficient machine learning methods<\/strong> and their use for the solution of applied engineering problems and industrial analytics.<\/p>\n In particular, we are working on regression based on Gaussian processes and kernel methods for multi-fidelity surrogate modeling and optimization, Deep Learning for 3D Data Analysis and manifold learning, on-line sequence learning for prediction and non-parametric anomaly detection.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The main goal of the lab is developing a novel multiscale modeling<\/strong> paradigm, where the emerging methods of Artificial Intelligence<\/strong> (AI) are incorporated into traditional mathematical modeling techniques, such as finite elements, Molecular Dynamics, Langevin Dynamics, Monte Carlo, and Lattice Boltzmann simulations.<\/p>\n We explore Complex Systems<\/strong>, such as multiphase Soft Matter<\/strong> (liquids, solutions, complex liquids, porous media, etc.) and Active Matter<\/strong> (self-driven particles, bacteria, animals, pedestrians, vehicles, robotic swarms, drones, etc.). Our primary focus is fundamental academic research that possesses a significant application potential and demand from industry and society.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The group researches the border of information theory, communications, and machine learning<\/strong> with a particular focus on the application of the research results in communications, including engineering and industry problems.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The AIDA Lab is dedicated to the study and application of Artificial Intelligence<\/strong> (AI) in dynamical environments<\/strong>, with a special focus on safety and trustworthiness<\/strong>. Among various machine learning approaches, we put a special focus on reinforcement learning<\/strong> – a methodology resembling the action of living beings in changing, uncertain environments that react by punishment and reward. Applied to the human economy, AI has to fulfill requirements on safety and trustworthiness, especially regarding personal data privacy. These requirements become particularly challenging in the dynamical application of AI, such as robotic, autonomous driving, medical therapy support, chemical engineering, energy management, etc.<\/p>\n The Lab seeks to apply an interdisciplinary approach<\/strong> by the fusion of machine learning<\/strong> with various fields, such as system and control theory, to develop novel dynamic AI methods<\/strong>.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The concept of neuromorphic computing<\/strong>, based on imitating and utilizing a biological nervous system’s basic principles, has evolved over the decades into an interdisciplinary field at the boundary between neuroscience and advanced computing. Nowadays, it is considered one of the most promising approaches for resolving the critical problems that conventional CMOS technology faces due to continual miniaturization and ever-increasing power consumption.<\/p>\n Owing to the low power performance and brain-inspired massively parallel computing principles, many bio-inspired algorithms and devices have been attempted in complex pattern recognition, image processing, and data mining. Intensive research has been conducted towards developing learning-based artificial synapses and neurons, attempting to reproduce the behavior of these two fundamental building blocks in biological neural networks.<\/p>\n With its focus on advancing frontiers of future neuromorphic computing technology, our team aims to develop\u00a0artificial synapses and neurons <\/strong>using various platforms, including memristors, which may be tailored to provide the element base for novel concept devices secure stable and controllable device operation.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The research group has a twofold focus:<\/p>\n The synergy between imaging systems and image interpretation is at the forefront of our scientific interest. The team members rely extensively on modern computer vision and deep learning methods<\/strong>, with frequent forays into adjacent disciplines for motivation and new ideas.<\/p>\n Historically, our most impactful discoveries belong to the area of biomedical imaging<\/strong>, with applications in optical microscopy, MRI, X-ray\/CT scanners<\/strong>, and others.\u00a0We have built two laboratories at Skoltech, where students can embed their computational algorithms into the consumer cameras and wearable sensors.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> Our research focuses on developing breakthrough numerical techniques<\/strong> for solving a broad range of high-dimensional problems<\/strong>.<\/p>\n The key ingredient is the effective decomposition of multidimensional arrays (tensors).\u00a0Our recent interests also involve graph mining, recommender systems, and topological shape\u00a0optimization.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> Interdisciplinary research topics of our group range from distributed deep learning<\/strong> to mathematical modeling<\/strong> of complex phenomena using modern supercomputers<\/strong>.<\/p>\n Using the CDISE flagship supercomputer \u201cZhores<\/a>\u201c, we solve challenging computational projects in fluid dynamics, plasma physics, photonics, and other areas using the hybrid approach: traditional parallel numerical modeling coupled with methods of machine learning. Additional research is devoted to evaluating modern high-performance computing (HPC) energy-efficient computing architectures.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> iMolecule\u00a0<\/strong>is an interdisciplinary research group committed to unlocking the full potential of artificial intelligence<\/strong> and machine learning<\/strong> for advanced molecular design<\/strong>. The research area lies at the intersection of molecular modeling, machine learning, and high-performance computing.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The research area of IoT Lab<\/strong> is focused on the creation of the IoT products and services<\/strong> for the digital transformation<\/strong> of industrial companies as well as the development of IoT-related solutions<\/strong> for the oil and gas sector, iron and steel industry, aviation, agriculture, etc.<\/p>\n In particular, we are currently equipped with testbeds for data collection, processing, and analysis from various enterprise data sources and elaborating the IoT platform suitable for further recommender systems deployment, predictive maintenance, and early production quality control.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The joint Skoltech-Sberbank laboratory works on the application of digital technologies<\/strong> in the industry<\/strong> and the development of Machine Learning algorithms<\/strong>.<\/p>\n In particular, we work in the fields of anomaly detection, kernel methods, embeddings, and Bayesian optimization. We hope that the development of principled methods in Machine learning will push forward both solutions in both theoretical and practical branches of data processing.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The Laboratory develops new fundamental approaches<\/strong> for training, testing, and storing parameters<\/strong> of deep neural networks<\/strong> based on tensor decomposition techniques<\/strong>. These approaches allow us to reduce by orders of magnitude computational complexity and required memory for the operation of the network while maintaining a high quality of prediction.<\/p>\n The Laboratory mission is to pursue cutting-edge research in the design and analysis of deep neural networks, tensor decompositions, tensor networks, and multiway analysis with many potential practical applications.<\/p>\n The Laboratory brings together several professors and young researchers in the fields of machine learning, computer vision, artificial intelligence, robotics, large-scale data analysis, mathematics as well as computational neuroscience, and bioinformatics.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> We are interested in applications of machine learning<\/strong> to bioinformatics<\/strong> and developing efficient algorithms for bioinformatics<\/strong>.<\/p>\n Examples of our topics: genotype-to-phenotype prediction, genome reconstruction from contaminated data, optimization of phylogenetic trees.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The Laboratory focuses on the instrumentation development<\/strong> and application of state of the art mass spectrometry techniques<\/strong>. All projects are closely related to:<\/p>\n Our primary goal is to develop mass spectrometry platforms and analytical solutions for a wide area of researches.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The laboratory is focused on robot navigation<\/strong>, and mainly our interests are within Path Planning and Perception<\/strong> and how to combine both into new solutions for robotics<\/strong>.<\/p>\n The\u00a0full autonomy\u00a0milestone for robots has not arrived yet, and we are slowly closing the gap; however, there are still many hindrances, such as uncertainty, scalability, and reliability.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div> The group focuses on inverse and ill-posed problems of mathematical physics<\/strong> and their applications in various fields of science and industry. Problems of this kind require research on modern methods of solution of integral and partial differential equations<\/strong>, regularization techniques<\/strong>, and high-performance computing<\/strong>.<\/p>\n We develop new algorithms and computational software for practical applications in medical physics (electro- and magnetoencephalography, tomography, etc.), geophysics (controlled-source electromagnetics and full-waveform inversion), various kinds of microscopy and tomography (such as electron microscopy and tomography in backscattered electrons), image and signal reconstruction. The group is also involved in the research on the application of deep learning techniques to ill-posed problems.<\/p>\n <\/div>\r\n <\/div>\r\n<\/div>
Prof.\u00a0Nikolay Brilliantov<\/a>,\r\nProf.\u00a0Alexey Vishnyakov<\/a>,\r\nProf.\u00a0Vladimir Palyulin<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Alexey Frolov<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof. Pavel Osinenko <\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof. Dmitry Yudin<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Dmitry Dylov<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t\n
Prof.\u00a0Ivan Oseledets<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Sergey Rykovanov<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Petr Popov<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Dmitry Lakontsev<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Dr.\u00a0Alexey Zaytsev<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Andrzej Cichocki<\/a>, Prof. Anh-Huy Phan<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Dmitry Yarotsky<\/a>,\r\nDr.\u00a0Gregory Kucherov<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Evgeny Nikolaev<\/a>,\r\nProf.\u00a0Yury Kostyukevich<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t\n
Prof.\u00a0Gonzalo Ferrer<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t
Prof.\u00a0Nikolay Koshev<\/a><\/span>\t\t\t<\/span>\r\n\t\t\t