[apologies for cross-posting]
Postdoc Modeling the Dynamics of Decision making and its Interplay with Memory
Robots in interactions with humans need to predict their behavior and safely navigate around them. Such predictions can be especially hard when humans exhibit sudden shifts in behavior, often under stress and time pressure. For instance, a pedestrian can
decide not to cross the road in front of an autonomous vehicle but then change their mind. Such sudden choice reversals could have particularly strong impact in high-stakes scenarios such as human-robot interactions in traffic. What are the cognitive mechanisms
of humans’ changes-of-mind in dynamic tasks under time pressure? How do our past experiences affect such decisions? How can we model the ‘stress factor’ mathematically? In this project, your aim will be to 1) develop a neurally plausible cognitive model to
explain changes-of-mind in dynamic decision-making and its interplay with memory, and 2) collect a large-scale online dataset on dynamic decision-making to evaluate the model. This proof-of-concept study will break through the fundamental limitations of the
current models and will establish the basis for real-life applications of neurally plausible decision-making models.
In this project, you will be able to develop your skills in mathematical modeling and numerical simulation. You will work at TU Delft in the Faculty of Mechanical Engineering jointly at the Department of Cognitive Robotics and the Delft Center for Systems
and Control, supervised by Dr. Matin Jafarian and Dr. Arkady Zgonnikov.
Requirements
Conditions of employment
More details/apply via
https://www.tudelft.nl/over-tu-delft/werken-bij-tu-delft/vacatures/details/?jobId=15583 until January 15, 2024.
Best regards,
Arkady Zgonnikov
Assistant Professor
Human-Robot Interaction Group
Department of Cognitive Robotics
Delft University of Technology