Instructors
Dr. Charley Wu (charley[dot]wu[at]uni-tuebingen[dot]de)
Dr. Charline Tessereau (charline[dot]tessereau[at]tuebingen[dot]mpg[dot]de)
Teaching Assistant
Hanqi Zhou (hanqi[dot]zhou[at]uni-tuebingen[dot]de)
Turan Orujlu (turan[dot]orujlu[at]uni-tuebingen[dot]de)
Mani Hamidi (manihamidi[at]gmail[dot]come)
General Information
Location: Ground floor seminar room, AI building, Maria-von-Linden-Str. 6, D-72076 Tübingen
Lecture: Thursdays from 10:00 - 12:00
Tutorial: Friday from 16:00 - 18:00
Office Hours: Charley (after the lecture); Charline (by appointment)
Course description and prerequisites:
The goal of the course is to teach modern approaches to machine learning in parallel with the science of human learning. Theories of human learning have directly inspired algorithms for training artificial agents, while advances in AI have provided novel insights into the nature of human intelligence. Machine learning algorithms are increasingly reaching or exceeding human-level expertise. Yet, they can still be fooled by rather trivial tasks that any human could solve. How have these two fields separately approached the study of artificial and biological intelligence? What are general principles of learning that are shared across human and machine learners?
The course covers how both biological and artificial systems i) learn from experience (Reinforcement Learning), ii) learn about concepts and categories, and ii) learn functions and programs. Each week, there will be one lecture and one tutorial session. No prerequisites are required, but some basic familiarity with linear algebra and basic mathematical notation will be necessary.
Grading
Note: We have revised the grading scheme due to several bus strikes impacting the ability of students to attend previous tutorials with pop quizzes. Pop quizzes will now be worth 20% of the grade, and we will currently use the best quiz grade out of the two remaining. In case of any future strikes, students who are able to attend and take the quiz will still be able to use that grade.
Schedule
Lecture: Thursdays from 10:00 - 12:00
Tutorial: Friday from 16:00 - 18:00
| Date | Remarks | Readings | Lecture | Lecturer | TA |
|---|---|---|---|---|---|
| Week 1: April 20,21 | Spicer & Sanborn (2019). What does the mind learn? | Introduction (slides) | Wu | Hamidi (tutorial slides) | |
| Week 2: April 27, 28 | [1] Behaviorism [2] What is a perceptron? (Blog post) | Origins of biological and artificial learning (slides) | Wu | Orujlu (tutorial slides) | |
| Week 3: May 4, 5 | Wikenheiser & Schoenbaum (2016) | Symbolic AI and Cognitive maps (slides) | Wu | Hamidi [Quiz 1] | |
| Week 4: May 11,12 | Dezfouli & Balleine (2012) | Introduction to RL (slides) | Tessereau | Orujlu (tutorial slides) | |
| Week 5: May 19 | Christihimmelfahrt | NA | No lecture | No tutorial | |
| Week 6: May 25, 26 | Botvinick, Niv, & Barto (2009) | Advances in RL (slides) | Tessereau | Orujlu [Quiz 2] | |
| Week 7: No classes | Pfinstpause | NA | No lecture | No Tutorial | |
| Week 8: June 9 | Fronleichnam | NA | No lecture | No Tutorial | |
| Week 9: June 15,16 | Murphy (2023) | Concepts and Categories (slides) | Wu | Hamidi (tutorial slides) | |
| Week 10: June 22,23 | Varma & Prasad | Supervised and Unsupervised learning (slides) | Wu | Zhou (programming challenge) | |
| Week 11: June 29,30 | Wu, Meder, & Schulz (in prep); sent via email | Function learning (slides) | Wu | Hamidi [Quiz 3] | |
| Week 12: July 6,7 | Botvinick et al., 2020 | Common tools for understanding brains and neural networks (slides) | Tessereau | Zhou (Barrett et al., 2019) (Botvinick et al., 2020) | |
| Week 13: July 13,14 | Landauer & Dumais (1997) | Language and semantics (slides) | Wu | Zhou [Quiz 4] | |
| Week 14: July 20,21 | Gershman (2023) | General Principles (slides) | Wu & Tessereau | Wu & Tessereau (tutorial doc) | |
| July 27, 10:30-12:00 | Exam 1 (same room as lectures) | ||||
| October 12, 10:30-12:00 | Exam 2 (4th floor seminar room, AI building) |