Trofimov Aleksandr Gennadievich — Doctor of Science, senior lecturer of the Department of Cybernetics at the Institute of Cyber Intelligent Systems of the National Research Nuclear University MEPhI.

Research advisor of the Laboratory of Neural networks technologies at the Institute of Cyber Intelligent Systems of the National Research Nuclear University MEPhI, member of the editorial board of the Journal Neurocomputers, member of program committee and organizing committee of the international conference "NEUROINFORMATICS".

University: National Research Nuclear University MEPhI (Moscow Engineering Physics Institute).

English proficiency: fluent.

Programme: Applied Mathematics and Information Science, Software Engineering.

Directions:

09.06.01 Computer Science and Computer Facilities (100)

List of Research Projects (participation/management):

• Russian Foundation for Basic Research, 15-29-01344, 2015-2017, participation.

Analysis of wave-like processes in the human brain.

• Russian Science Foundation, 14-28-00234, 2015–2017, participation.

In search of the "I": Interdisciplinary research on the initiation of voluntary action.

List of possible research topics:

1. Development of an intelligence system for image geotagging.
2. Vulnerability study of existing types of graphic captcha as a cyber security elements. Development of a computer-aided system for captcha recognition and generation of hack-resistant captcha types.
3. Research of the success of start-ups based on open source text analysis, including social networks.
4. Development of a predictive analytics system for early detection of breakdowns at power plants.
5. Development of methods for detecting adversarial attacks on ML models and methods of protection against them.

Field of Study: Machine learning and neural networks.

Research interests:

1. Image processing and computer vision.
2. Analysis of natural language texts.
3. Predictive analytics.
4. Machine learning models security.

Key features of the programme: Opportunity to work with MEPhI graphics accelerators of the Tensor computing module (1280 GPU tensor cores, 10240 CUDA cores).

Specified requirements for a candidate:

• Knowledge of mathematical analysis, linear algebra, probability theory and mathematical statistics
• Python and MATLAB skills;
• Knowledge of Python scientific programming libraries, including scikit-learn.

Number of publications in Web of Science and Scopus (9 publications):

• Trofimov A.G., Bogatyreva A.A. A method of choosing a pre-trained convolutional neural network for transfer learning in image classification problems // Studies in Computational Intelligence, 2020, Vol. 856 pp. 263-270.
• Trofimov A.G., Kuznetsova K.E., Korshikova A.A. Abnormal Operation Detection in Heat Power Plant Using Ensemble of Binary Classifiers //  Advances in Neural Computation, Machine Learning, and Cognitive Research, 2019, V.799, Pp.227-233.
• Trofimov A.G., Shishkin S.L., Kozyrskiy B.L., Velichkovsky B.M. A Greedy Feature Selection Algorithm for Brain-Computer Interface Classification Committees // Procedia Computer Science, 2018, V.123, Pp. 488-493.
• Trofimov A.G., Velichkovsky B.M., Shishkin S.L. An Approach to Use Convolutional Neural Network Features in Eye-Brain-Computer-Interface // Studies in Computational Intelligence, 2018, V.736, Pp.132-137.
• Shishkin, Y. Nuzhdin, E. Svirin, A. Trofimov, A. Fedorova, B. Kozyrskiy, B. Velichkovsky. EEG Negativity in Fixations Used for Gaze-Based Control: Toward Converting Intentions into Actions with an Eye-Brain-Computer Interface // Frontiers in Neuroscience. 2016, V.10, P. 1-20.