< Back to previous page


Probabilistic programs for dynamic relational worlds

In order to represent complex real-world environments, which typically involve multiple entities as well as the relationships that hold amongst them, it is necessary to employ both logical and relational representations. Furthermore, considering the inherent uncertainty in these environments, one needs models that cannot only cope with dynamic relational states but also with the uncertainty.

The overall goal of this project is thus to contribute novel probabilistic programming and statistical relational artificial intelligence techniques that allow one to model, reason and learn about dynamic relational worlds, cope with the inherent uncertainty and that are applicable to real-world tasks in game playing and cognitive robotics.

Date:4 Jan 2021  →  Today
Keywords:Artificial Intelligence, Machine Learning, Logic, Logic Programming, Neural Networks, Deep Learning
Disciplines:Knowledge representation and reasoning, Neural, evolutionary and fuzzy computation
Project type:PhD project