Mikhail received his Bachelor and Master degrees in Applied Mathematics and Physics from the Moscow Institute of Physics and Technology in 2010. His Master thesis addressed various aspects of training shallow neural networks, including parameters initialization, automatic selection of network size and a special optimization algorithm for accelerated learning.
After graduation, Mikhail applied the developed method to various problems in aerospace engineering as Researcher at the Institute for Information Transmission Problems. In 2011, Mikhail joined DATADVANCE, a resident company of Skolkovo Innovation Center, as Senior Researcher. Mikhail developed and implemented several machine learning algorithms in the context of an industrial data analysis library intended primarily for aerospace and automotive. He also accomplished several industrial data analysis projects, including modeling of a spacecraft aerodynamics, building a predictive model for clogging of a nuclear power plant cooling system (both as a project leader), improvements in a design of a Formula 1 car side panel. In 2015, Mikhail went back to IITP, proved asymptotical optimality of a developed machine learning method for structured data and defended his Candidate of Science thesis.
In 2015, Mikhail headed a small research group at IITP, which became an individual scientific department of IITP in January 2016. The team contributes to data analysis in neuroscience focusing its research on the analysis of neuroimaging data (various modality of MRI), algorithms for EEG analysis and mining of large databases of clinical information related to neurodegenerative diseases.
Development of new machine learning algorithms for structured data, employing new machine learning algorithms for analysis of structured neuroimaging data, including mass spectrometry data and different modality of magnetic resonance imaging data.