cv
Table of contents
- General Information
- Education
- Teaching Experience
- Research Experience
- Honors and Awards
- Academic Interests
- Other Interests
General Information
Full Name | Aleksandr Panov |
Place of work | Moscow, Russia |
Languages | Russian, English |
Education
-
2015 PhD
Institute for System Analysis of RAS, Moscow, Russia - Specialized in modeling of goal-oriented behavior of intelligent agents and their coalitions
- Thesis title is Investigation of methods, development of models and algorithms for formation of elements of sign-based worldview of the actor
-
2011 Master’s degree
Moscow Institute of Physics and Technology, Department of Applied Mathematics and Management, Dolgoprudny, Russia - Majors are technologies of active databases, computer graphics, game theory and decision making, effective algorithms, decomposition in optimization
- Specialized in logical methods (AQ, JSM) of data mining and multi-agent systems
- Thesis title is Investigation and modeling of group behavior for multifunctional agents
-
2009 Bachelor of Physics
Novosibirsk State University, Department of Physics, Novosibirsk, Russia - Majors are operational systems, digital integrated circuits, introduction to CAD, microprocessors, information networks and systems, object-oriented analysis and design
- Specialized in semantic integration of databases.
- Thesis title is Semantic integration of biological databases
Teaching Experience
-
2019 - Present Head of AI Master Program
Moscow Institute of Physics and Technology, Phystech School of Applied Mathematics and Informatics, Dolgoprudny, Russia - Seminars on Basis of Operation Systems and Basis of Object-Oriented Programming
- Lectures on Introduction in AI and Reinforcement Learning
-
2015 - 2019 Associate Professor
National Research University Higher School of Economics, Faculty of Computer Science, Moscow, Russia - Seminar on Intelligent Data Mining
-
2011 - 2016 Assistant Lecturer
Peoples’ Friendship University of Russia, Department of Computer Science, Moscow, Russia - Lectures on Intelligent Dynamic Systems, Theoretical Computer Science and Intelligent Data Analysis
Research Experience
-
2021 - Present Principal Research Fellow
Artificial Intelligence Research Institute, Neural-symbolic team, Moscow, Russia - Leading non-profit organization in the field of Artificial Intelligence – http://airi.net
- Reinforcement learning in multi-agent systems
- Switching algorithms of planning-based and learning-based multi-agent path finding methods
- Monte-Carlo approach in multi-agent systems
- Neural-symbolic integration
- Disentangled representations and object-oriented world models
- Vector symbolic architectures in VQA and robot navigation setting
-
2010 - Present Principal Research Fellow, Head of Laboratory
Federal Research Center “Computer Science and Control” of Russian Academy of Sciences, Moscow, Russia - Leading academic institute in Computer Science and High-performance computing – http://frccsc.ru
- Cognitive modeling
- Psychologically inspired models of human behavior based on theory of sign-based world model
- Biologically inspired models of sign components - image, significance and personal meaning
- Algorithms of behavior planning and goal setting procedures
- Machine learning and multi-agent systems
- The composite logical method to extract cause-effect relationships
- Algorithms of planning and role distribution in coalition of cognitive agents
- Cognitive Robotics
- Multi-layer control system for coalition of cognitive robots
-
2018 - Present Director, Head of Laboratory
Moscow Institute of Physics and Technology, Center for Cognitive Modeling, Dolgoprudny, Russia - Leading University in Russia in Physics and Computer Science – http://cogmodel.mipt.ru
- Applied research in self-driving cars and mobile robotics
- New framework for behavior planning of self-driven cars based on Apollo-auto
- Original methods of neural-based object segmentation, detection, tracking for mobile robots
- Reinforcement learning
- Hierarchical reinforcement learning and learning from demonstrations
- Learning-based methods for visual navigation in indoor scenes
- Neuromorphic computing
- Architecture of the hierarchical intrinsically motivated agent (HIMA)
- Improved variants of hierarchical temporal memory
-
2015 - 2018 Research Fellow
National Research University Higher School of Economics, Laboratory of Process-Aware Information Systems (PAIS Lab), Moscow, Russia - Leading University in Russia in Economics and Computer Science - http://hse.ru
- {"title"=>"Investigation of learning mechanisms based on sign representations in the problem of collective behavior planning"}
Honors and Awards
-
2023 - I led the SkillFusion team, which secured the first place in the CVPR Habitat competition
-
2019 - I led the CDS team, which took first place in the NeurIPS MineRL competition
-
2017 - Russian Academy of Sciences Medal for Young Scientists
Academic Interests
-
Computer Cognitive Modeling
- Semiotics
- Sign-based World Model
-
Embodied AI
- Cognitive Robotics
- Behaviour planning
- Language models for planning
-
Multi-agent systems
- Multi-agent planning
- Multi-agent reinforcement learning
-
Reinforcement Learning
- Model-based learning and planning
- Transformer-based world models
- Structured world models
Other Interests
- Hobbies: snowboarding, ski-touring, badminton, football