Sept 6 - Sept 11, 2021
Cognitive modeling has become ubiquitous for investigating cognitive processes in Psychology and their natural representations in Neuroscience. However, mastering the first step in the field of computational cognitive neuroscience can be a bit overwhelming for interested scientists just entering the field.
The Cognitive Modeling Academy Hamburg aims to ease these initial steps. Our goal is to provide the theoretical background of cognitive modeling in an accessible way. We will guide you through the necessary step of a cognitive modeling study and provide practical exercises along the way for early hands-on experience. Participants will apply these new skills in a week-long research project that they will work on in small groups.
We will present in detail two prominent and highly successful modeling frameworks in this course: reinforcement learning and sequential sampling models (eg drift diffusion models). Reinforcement learning is a general computational model for learning about rewards a complex environment and for planning how to obtain them. Sequential sampling models provide a more fine-grained look at the core decision process and provides not only a theoretical framework for which decisions are made, but also how long it takes to make them.
Defining a cognitive model
Estimating free model parameters (model fitting)
Choosing the best model (model selection)
Predicting new data (model validation)
Reward expectations and learning from errors
RL models for learning associations
RL models for choosing between different options
RL models for decisions in complex environments
Dynamic of decisions
Decisions as evidence accumulation
Drift diffusion models
Combining RL and SSM
Frequently asked questions
The Cognitive Modeling Academy Hamburg is an introductory course targeted at researchers (MS, PhD, Postdoc, Professors) without or very limited experience in cognitive modeling that are interested in learning about it. We will cover the basic steps of a cognitive modeling study and introductions to reinforcement learning and sequential sampling models (e.g. drift diffusion model) (see About). If you conducted a modeling study or have experience in modeling behavioral data, then this course may be too easy for you, although a systematic introduction can still be helpful in organizing the research and understanding the underlying concepts.
The Cognitive Modeling Academy Hamburg will be held this year as an online event on gather.town from Sept 6-11, 2021 (see Schedule).
Thanks to our sponsor VolkswagenStiftung, we are able to offer the course at a reduced rate of 250 €. In addition, we can offer up to stipends to 4 students, who can participate for free (see Financial Support).
If you want to get the most out of this course and participate in the exercises and student research projects, then you will need basic knowledge in the R programming language. Unfortunately, this cannot be covered during the course. On the Resource page we have provided links to online R and modeling tutorials.
Contact us via email firstname.lastname@example.org