Decision-making under uncertainty in partially observable domains is at the heart of any single- and multi-agent autonomous system acting with imperfect information in stochastic environments. In this project we develop novel approaches for simplification of these challenging problems, aiming to accelarate decision-making while providing formal performance guarantees. The project seeks to advance fundamental theory as well as develop prototype demonstrations on real robots.
Prerequisites:
- Strong analytical skills, passion for rigorous mathematical formulations
- Strong programming skills (preferably Python or Julia).
- Background in probability and measure theory, MDPs/POMDPs, and (deep) reinforcement learning is an advantage.
Academic supervisor:
- Prof. Vadim Indelman (email)
Duration: 1 or 2 semesters