Skolkovo Institute of Science and Technology
Computational Intelligence Group
Our research focuses on developing breakthrough numerical techniques for solving a broad range of high-dimensional problems. The key ingredient is the effective decomposition of multidimensional arrays (tensors). Our recent interests also involve graph mining, recommender systems and topological shape optimization.
deliverables
We will provide new insights into the theory of AI and based on it develop more accurate, robust and fast models for different applications
improving existing models, including GANs, unsupervised and representation learning
construction of methods of representation learning in the small-data regime
construction of new architectures of deep neural networks
dataset dimensionality reduction
study of robustness and expressive power of deep neural network architectures
physical-based machine learning and data-driven modelling of real world systems
recommender systems
contacts