Associate Research Professor at Institute for Cybernetics of NASU
Dmitriy has graduated from the Moscow Institute of Physics and Technology, Faculty of Applied Mathematics. Candidate of Physico-Mathematical Sciences, Doctor of Applied Mathematics at the Paul Sabatier University, Toulouse (France). Coordinator of the prospective neurobiological project Rybka Project. Also he has worked at the universities of Toulouse and Grenoble (France), University of Massachusetts and Harvard (USA).
Topic of presentation: State representation learning and associative memory for intelligent agents
The main points of the presentation:
Most real-world world tasks are hopelessly complex from the point of view of reinforcement learning mechanisms.
In this talk we will give an overview of methods for finding efficient state representations from agent’s observations. We will consider Bayesian inference with infinite capacity models, and explore links representation learning and the machine learning problem of transfer learning. Then we will present models of neural associative memory and their interplay with machine learning.