About

My name is Georgii “George” Nigmatulin, I’m a Machine Learning Research Engineer at Samsung R&D Institute Russia, creating new AI features for smartwatches based on sensor time-series signals. In July 2022, I received MSc in Applied Mathematics and Physics after finishing my studies at the Moscow Institute of Physics and Technology.

My research interests include applying machine (deep) learning, causal and bayesian methods, reinforcement learning to problems in the field of quantitative finance, particularly to portfolio optimization, market dependency structure, factor investing and predictive analytics. I’m also drawn to specific areas like graph neural networks for time series networks, frequent patterns mining, regime-switch modeling, stochastic control and optimization, online portfolio selection.

My research experience, which I’m currently engaged in at Samsung, concerns proceeding signals by ML algorithms, dependency structure, end-to-end development of new features. In academic research, I’ve worked in the fields of Uncertainty quantification, Anomaly detection, Deep Learning and Active learning in the Skoltech-Sberbank joint laboratory. Previously at top-1 regional bank, I was engaged in predictive modeling of withdrawals from ATM networks and created a discrete optimization system that controls cash management.