Only open to MSc Psychology (research) students.
Reinforcement learning is the adaptive process by which we learn to predict the consequences of our behavior through interactions with our environment. Over the last few decades, this has been intensively studied in a range of fields including psychology, artificial intelligence, animal and human neuroscience and economics. A related but largely separate literature is concerned with how we make optimal decisions based on noisy sensory information. Computational models of the putative underlying mechanisms of learning and decision making play a central role in both of these research fields. This course is intended to provide an overview of the computational and neural mechanisms of reinforcement learning and (value-based and perceptual) decision making, and to gain insight in the use of computational models to account for experimental data. Topics include classical and instrumental conditioning, Markov decision processes, the exploitation-exploration tradeoff, and sequential sampling models of perceptual decision-making.
The course meetings will be based on empirical papers that have made a significant contribution to the field and on papers that review a substantial body of research. Students will gain hands-on experience with programming computational models and fitting them to experimental data. Specifically, each student will implement both a reinforcement-learning model and a decision-making model, one in a guided manner in class, and another one individually. They will present the results of their models in a scientific poster. Students will also write a short essay about a current issue in the reinforcement learning or decision-making literature.
Upon completion of the course, students will be able to:
Explain the key paradigms and models currently employed in the fields of reinforcement learning and decision making (based on a theoretical overview provided by the lectures);
Understand and program computational models, and fit them to experimental data. This ability will be a crucial tool for students due to the fast-growing demand for skills related to programming and computational modelling in the job market in- and outside of academia;
Identify the steps necessary to translate abstract theoretical concepts into concrete experiments;
Account for experimental findings such as differences in reaction times and accuracy through modeling;
Critically discuss current issue in the reinforcement learning and/or decision making literature.
For the timetables of your lectures, work groups and exams, please select your study programme in:
Students need to enroll for lectures and work group sessions.
Master’s course registration
Students are not automatically enrolled for an examination. They can register via uSis from 100 to 10 calendar days before the date. Students who are not registered will not be permitted to take the examination.
Registering for exams
Mode of instruction
8 2-hour seminars
The assessment of the course is based on:
50% poster presentation.
The Institute of Psychology follows the policy of the Faculty of Social and Behavioural Sciences to systematically check student papers for plagiarism with the help of software. Disciplinary measures will be taken when fraud is detected. Students are expected to be familiar with and understand the implications of this fraud policy.
O’Reilly, J.X. (2013). Making predictions in a changing world—inference, uncertainty, and learning. Frontiers in Neuroscience, 7:105
Cohen, J. D., McClure, S. M., & Yu, A. J. (2007). Should I stay or should I go? How the human brain manages the trade-off between exploitation and exploration. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 362(1481), 933-942.
Dayan, P., & Daw, N. D. (2008). Decision theory, reinforcement learning, and the brain. Cognitive, Affective & Behavioral Neuroscience, 8(4), 429-453.
Voss A, Nagler M, Lerche V (2013). Diffusion models in experimental psychology: a practical introduction. Experimental Psychology, 60(6), 385-402.
Franka Richter firstname.lastname@example.org