Single-Course Engelsk 5 ECTS

Cognitive Modelling

DTU Compute

Overall Course Objectives

In this course you will learn about probabilistic models of cognitive processes. You will gain proficiency in analysing data from behavioral psychophysical experiments using signal detection theory and the psychometric function. We will analyse these standard methods and realise that they rooted in underlying models and assumptions. Subsequently, we will see that these models can be fit into a broader Bayesian modeling framework, which we will use to build custom probabilistic models of cognitive processes.

The course will also delve into linear models of neural encoding of information and how we can use these models to decode neural representations from behavioral data.

A significant focus of this course lies in mastering the fitting and evaluation of customised probabilistic models using the maximum likelihood principle and numerical optimisation. Proficiency in coding, particularly in Python or MATLAB, is expected from students. Additionally, a foundational understanding of probability theory is important. Throughout the course, emphasis will be placed on the ability to construct custom models, minimising reliance on pre-existing functions and toolboxes.

See course description in Danish

Learning Objectives

  • Analyse data from psychophysical data using signal detection theory and the psychometric function
  • Analyse perception from a Bayesian perspective
  • Analyse linear models of neural coding
  • Describe the relationship between neural and cognitive models
  • Fit a custom probabilistic model of a cognitive process to behavioural data
  • Evaluate custom probabilistic model using cross-validation
  • Simulate data from a probabilistic model
  • Use simulated data to evaluate model and parameter recovery
  • Design experiments to test probabilistic models of cognitive processes
  • Discuss the relation between qualitative and quantitative models of cognitive processes

Course Content

The topics of the course are introduced in weekly lectures. Great emphasis is placed on practical work where students gain proficiency in fitting and evaluating custom probabilistic models of cognitive processes through project work and exercises.

The course does not include mandatory assignments, and participation in course activities is strongly recommended, but optional. Nonetheless, students will have the chance to receive feedback on their project work and exercises.

Recommended prerequisites

02450/02454/02464, Probability theory, fundamental machine learning and basic programming. Knowledge or interest in cognitive science.

Teaching Method

Lectures, exercises and project work

Faculty

Tobias Søren Andersen

Contact or read more about Tobias Søren
Tobias Søren Andersen

See course in the course database.

Registration

Language

Engelsk
Duration

13 weeks
Institute

Compute
Place

DTU Lyngby Campus
Course code 02458
Course type Candidate
Semester start Week 36
Semester end Week 49
Days Thurs 8-12
Price

9.250,00 DKK

Related educations

Registration