This seminar will cover different approaches to synthesize models for agent behavior, divided into two tracks: The first track focuses on creating behavior models using Learning-from-Demonstration and Behavior Trees. The second track investigates approaches from the field of Neurosymbolic AI to synthesize domain- and action models.
This course is about the logic of knowledge bases, in two distinct but related senses. On the one hand, a knowledge base is a collection of sentences in a representation language that entails a certain picture of the world represented. On the other hand, having a knowledge base entails being in a certain state of knowledge where a number of other epistemic properties hold. One of the principal aims of this course is to develop a detailed account of the relationship between symbolic representations of knowledge and abstract states of knowledge. Students wishing to attend the course should be familiar with first-order predicate logic.
The lecture starts on Friday, April 21, 2023.
The proseminar will be on different (sub-)topics from artificial intelligence. We largely follow the lines of the well known textbook by Stuart Russell and Peter Norvig “Artificial Intelligence - A Modern Approach”.