Aleksandr Shevchenko

IST Austria
Am Campus 1
3400 Klosterneuburg, Austria

email: alex.shevchenko@ist.ac.at

Supervisor
Dan Alistarh
(VGSCO and IST), Marco Mondelli (IST)

 

 

Research Interests

Theoretical Foundations of Deep Learning

Scientific CV

Education

Since 09/2019: PhD student in Mathematics & Computer Science, Institute of Science and Technology Austria, Klosterneuburg
09/2018 – 09/2019: Master student in Statistical Learning Theory, Higher School of Economics/Skoltech, Moscow, Russia
09/2014 – 08/2018: BSc in Computer Science with distinction, Higher School of Economics, Applied Mathematics and Informatics, ML specialization, Moscow, Russia

Work Experience

09/2017 – 09/2019: Bayesian Methods Research Group: Research assistant, supervised by Anton Osokin and Dmitry Vetrov, Moscow, Russia
11/2018 – 09/2019: Samsung-HSE laboratory: Big Data and Information Retrieval School, Higher School of Economics, Research Assistant, Moscow, Russia
04/2018 – 11/2018: Center of Deep Learning and Bayesian Methods,
Big Data and Information Retrieval School, Higher School of Economics, Research Assistant, Moscow, Russia
07/2017 – 11/2017: IPONWEB, Intern - Junior Analyst, R&D Team, Moscow, Russia
02/2017 – 07/2017: DPL Lab, ML Engineer, Counter Recognition Team, Moscow, Russia

Teaching Experience

09/2017 – 06/2018: National Research University Higher School of Economics, Teaching Assistant for Machine Learning, Moscow, Russia
02/2017 – 06/2017: National Research University Higher School of Economics, Teaching Assistant for Mathematical Analysis, Moscow, Russia

Publications

Accepted/Published

Landscape Connectivity and Dropout Stability of SGD Solutison for Over-parameterized Neural Networks
A Shevchenko, M Mondelli
37th International Conference on Machine Learning (ICML)