News
Guest Lecture on LLMs in Symbolic PlanningWritten on 20.10.25 by Joerg Hoffmann Hi all, for those of you interested in the use of LLMs in planning, we are happy to announce that there will be a guest lecture on this topic on January 22. The lecture will be given by Katharina Stein, who is doing her PhD work on this topic (among others). best regards, Jörg… Read more Hi all, for those of you interested in the use of LLMs in planning, we are happy to announce that there will be a guest lecture on this topic on January 22. The lecture will be given by Katharina Stein, who is doing her PhD work on this topic (among others). best regards, Jörg Hoffmann ps. The content of this guest lecture will not be exam relevant, it's just for interested students.
|
The first lecture will be on October 14th, 14:00 in E1 3, Lecture Hall 002. See also the Timetable.
Course Description
This course offers an introduction to the field of Artificial Intelligence (AI). We focus on the theory and practice of sequential decision making where the AI agent needs to decide about actions in a complex environment so as to achieve a long-term objective such as reaching a goal, maximizing expected reward, or winning a game. For example, games like Go and Chess are sequential decision making problems. The course includes aspects of both, symbolic and sub-symbolic AI (aka data-driven AI aka ML), as well as combinations thereof, and therewith provides students with a glimpse of one of the most important challenges in AI today -- combining symbolic and sub-symbolic AI. A prominent example of an AI system leveraging such a combination is the AlphaGo/Zero system series, which changed the world of computer game-playing, and whose ingredients and architecture we will understand as part of the course.
Students learn to master techniques developed in the fields of search algorithms, classical planning, Markov decision processes, and game playing. The lecture is accompanied by C++ programming projects in which students implement some of the concepts and algorithms encountered in the lecture. With the knowledge acquired in this course, students are knowledgeable in crucial aspects of AI, and are well-prepared for student assistants jobs as well as BSc and MSc theses at FAI and other AI-related research groups in Saarbrücken.
For more information, please check our organization page.
Prerequisites. Solid knowledge in algorithms and data structures is necessary to follow this course. Solid knowledge in imperative programming is also necessary. In particular, the projects require to program in C++ and (some) Python. Basic knowledge in machine learning will help, but is not strictly required; the same goes for basic knowledge in complexity theory.
Note. Students who have passed the Artificial Intelligence course in previous years are not allowed to retake the exam.
The first lecture will primarily address organizational matters. We aim to answer any remaining questions there. If for some reason you cannot wait until then, you may contact one of the teaching assistants directly or visit the forum.