“I worked under Mikhail’s lead on AI project, based on deep learning and neural networks lab, at Moscow Institute of Physics and Technology. Mikhail has a very deep understanding of machine learning and artificial intelligence development process. His project management skills are high, he has the abilities to push big, complicated, cross scientific IT project forward. Mikhail is certainly very professional project’s head and I can recommend him for any future employer, or other partnerships.”
About
Experience & Education
Volunteer Experience
-
Author of idea and chief organizer
Moscow Institute of Physics and Technology (State University) - MIPT, Phystech
- Present 9 years 1 month
Science and Technology
The first Russian deep learning summer school-hackathon DeepHack.Game — a week-long hackathon to improve
DeepMind code for playing Atari games (see http://game.deephack.me). It was a big success (see
http://www.prweb.com/releases/2015/07/prweb12876261.htm). The academic, free for participants but a competitive event combined hacking with a crash course of educational lectures by leading
researchers in the field including Yoshua Bengio, Ruslan Salakhutdinov, Jürgen Schmidhuber and…The first Russian deep learning summer school-hackathon DeepHack.Game — a week-long hackathon to improve
DeepMind code for playing Atari games (see http://game.deephack.me). It was a big success (see
http://www.prweb.com/releases/2015/07/prweb12876261.htm). The academic, free for participants but a competitive event combined hacking with a crash course of educational lectures by leading
researchers in the field including Yoshua Bengio, Ruslan Salakhutdinov, Jürgen Schmidhuber and others. The winning team wrote a scientific paper based on their results and approaches. The paper is accepted to a workshop at the Neural Information Processing Systems (NIPS) conference.
Publications
-
DeepPavlov: Open-Source Library for Dialogue Systems
Proceedings of ACL 2018, System Demonstrations, 122-127
-
The First Conversational Intelligence Challenge
The NIPS'17 Competition: Building Intelligent Systems, 25-46
-
ConvAI Dataset of Topic-Oriented Human-to-Chatbot Dialogues
The NIPS'17 Competition: Building Intelligent Systems, 47-57
-
Episodic memory transfer for multi-task reinforcement learning
Biologically inspired cognitive architectures 26, 91-95
-
The 2nd International Workshop on Search-Oriented Conversational AI
Proceedings of the 2018 EMNLP Workshop 2nd SCAI
-
Patterns of spiking activity of neuronal networks in vitro as memory traces
Biologically Inspired Cognitive Architectures (BICA) for Young Scientists pp 241-247
-
Simulation of Learning in Neuronal Culture
Biologically Inspired Cognitive Architectures (BICA) for Young Scientists, 47-52
-
NARLE: Neurocognitive architecture for the autonomous task recognition, learning, and execution
Biologically Inspired Cognitive Architectures, v. 13, pp. 91-104
-
Simulation of spontaneous activity in neuronal cultures with long-term plasticity
Matematicheskaya Biologiya i Bioinformatika 10 (1), 234-244
-
Evolution of event and delay controlled neuronal network for locomotion
In proc.: Int’l Conf. Genetic and Evolutionary Methods GEM’14. – С. 41-47
-
Evolution, development and learning with predictor neural networks
In proc.: ALIFE 14: The Fourteenth Conference on the Synthesis and Simulation of Living Systems. – Т. 14. – С. 457-464
-
Neuroevolution of sequential behavior in multi-goal navigation task
In proc.: ALIFE 14: The Fourteenth Conference on the Synthesis and Simulation of Living Systems. – Т. 14. – С. 771-777
-
Neuroevolution results in emergence of short-term memory in multi-goal environment
Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference GECCO ’13. New York, NY, USA: ACM, С. 703–710
-
Adaptive learning through variation and selection
In Seel, Norbert M. (Ed.) Encyclopedia of the Sciences of Learning, vol.1, p. 116-118.
-
Adaptive functional systems: Learning with chaos
Chaos 20(4), 045119.
Other authors -
-
Basic principles of adaptive learning through variation and selection
In Bullock S. et al. (eds.) Artificial Life XI, pp. 88-93. MIT Press, Cambridge, MA.
-
Simulation of the Evolution of Aging: Effects of Aggression and Kin-Recognition
In F. Almeida e Costa et al. (eds.): ECAL 2007, Lecture Notes in Computer Science 4648, pp. 84-92.
Other authors -
-
Evolution of cooperative strategies from first principles
Nature 440, pp. 1041-1044.
Other authors -
-
Artificial Life Meets Anthropology: A Case of Aggression in Primitive Societies
In Capcarrere M. et al. (Eds.): ECAL 2005, Lecture Notes in Computer Science 3630, pp. 655 – 664.
-
Theory of Functional Systems, Adaptive Critics and Neural Networks
In Proceedings of IEEE International Joint Conference on Neural Networks 2004, pp. 1787-1792.
Other authors -
-
Tracking the Trajectories of Evolution
Artificial Life 10(4), pp. 397-411.
-
Measuring the Dynamics of Artificial Evolution
Banzhaf W. et al. (Eds.): ECAL 2003, Lecture Notes in Computer Science 2801, pp. 580-587.
-
Functional Systems Network Outperforms Q-learning in Stochastic Environment
Procedia Computer Science 88, 397-402
-
Project "Animat Brain": Designing the Animat Control System on the Basis of the Functional Systems Theory
In Butz M.V. et al. (Eds.): Anticipatory Behavior in Adaptive Learning Systems, Lecture Notes in Artificial Intelligence 4520, pp. 94-107.
Other authors -
Projects
-
DeepPavlov
- Present
An open-source library DeepPavlov is tailored for development of conversational agents. The library prioritises efficiency, modularity, and extensibility with the goal to make it easier to develop dialogue systems from scratch and with limited data available.
DeepPavlov is designed for:
* development of production-ready chatbots and complex conversational systems;
* research in dialogue systems and NLP in general.
The framework provides conversational AI application…An open-source library DeepPavlov is tailored for development of conversational agents. The library prioritises efficiency, modularity, and extensibility with the goal to make it easier to develop dialogue systems from scratch and with limited data available.
DeepPavlov is designed for:
* development of production-ready chatbots and complex conversational systems;
* research in dialogue systems and NLP in general.
The framework provides conversational AI application developers and researchers with:
* a set of pre-trained NLP models, pre-defined dialogue system components (ML/DL/Rule- based) and pipeline templates;
* a framework for implementing and testing their own dialogue models;
tools for integration of applications with existing infrastructure (messengers, helpdesk software etc.);
* a benchmarking environment for conversational models and uniform access to relevant datasets.
The library features a wide range of state-of-the-art solutions for NLP tasks which are used in dialogue systems. These NLP functions address low-level tasks such as tokenisation and spell-checking as well as a more complex tasks including recognition of user intents and entities. They are implemented as modules with unified structures and are easily combined into a pipeline. The library provides a rich set of pre-trained models for a cold start. A user may start with a model nearest to her task, adapting and fine-tuning it to optimise performance.
Unlike many other frameworks, DeepPavlov allows the combination of trainable components with rule-based components and neural networks with non-neural ML methods. In addition to that, it allows end-to-end training for a pipeline of neural models.
Languages
-
English
Full professional proficiency
-
Russian
Native or bilingual proficiency
Recommendations received
1 person has recommended Mikhail
Join now to viewOther similar profiles
-
Daniel Kornev
Connect -
Dilyara Zharikova (Baymurzina)
PhD, Researcher, CTO at Sentius
Connect -
Aydar Bulatov
Deep Learning Researcher – DeepPavlov.ai
Connect -
Alla Chepurova
ML engineer
Connect -
Nick Pachnev
Connect -
Victor Lempitsky
Connect -
Ivan Maksimov
NLP Engineer
Connect -
Tatiana Shavrina, PhD
Connect -
Artem Shelmanov
NLP research scientist
Connect -
Brett Brewer
Connect
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Mikhail Burtsev
-
Mikhail Burtsev
Junior Front-end developer. React the reality!
-
Mikhail Burtsev
-
Mikhail Burtsev
--
-
Mikhail Burtsev
--
6 others named Mikhail Burtsev are on LinkedIn
See others named Mikhail Burtsev