Steven Miletić's research focuses on cognitive and neural modelling of decision and learning processes in humans, with emphasis on the human subcortex and on advanced methodology including hierarchical Bayesian approaches and ultra-high field fMRI. .

More information about Steven Miletic

Steven Miletić's research focuses on cognitive and neural modelling of decision and learning processes in humans, with emphasis on the human subcortex and on advanced methodology including hierarchical Bayesian approaches and ultra-high field fMRI.

Research

Miletić's research aims to understand the cognitive and neural processes that underlie human decision making and error-driven learning. Through advanced modelling of behavioural data (response time distributions, accuracy, learning patterns) and neural data (BOLD responses measured with ultra-high field fMRI) of decision making tasks, the ultimate goal is to obtain a joint understanding of the behavioural and neural processes that underlie decision making. Typical modelling frameworks include evidence accumulation models (EAMs; e.g., the diffusion decision model and racing diffusion models), reinforcement learning models, and Bayesian estimation approaches. On the neural side, his work puts a strong emphasis on the subcortical contributions to human behavior, including the basal ganglia and dopaminergic midbrain.

Recent projects include studying and modelling sequential patterns in choice behaviour; studying vasculature, the development of joint BOLD-fMRI-EAM modelling techniques; and combining reinforcement learning and evidence accumulation models. He also continuously works on methods to improve signal quality of 7 Tesla fMRI in the human subcortex.

Teaching

Miletić teaches the courses Cognitive Neuroscience, Introduction to Cognitive Science, Cognition and Computation and Neurochemistry.

Short CV

Miletić obtained his PhD (cum laude) at the integrative model-based cognitive neuroscience group (IMCN) at the University of Amsterdam. His dissertation focused on modelling structure and function of the human subcortex. The human subcortex comprises approximately a quarter of the brain volume, but is systematically underrepresented in studies in cognitive neuroscience. His thesis argued that this is, for a major part, a measurement problem, and he subsequently developed methodology using 7 Tesla (f)MRI and advanced modelling to study the human subcortex. In his subsequent post-doc in the same group, he focused on the development of joint modelling techniques, on improving our understanding of the vasculature in the subcortex, and became interested in the sequential structures in behavioral data in decision making, which he believes can provide a new window into cognitive processes. In 2023, he became assistant professor at the cognitive psychology department of Leiden University.

His full CV can be found here.

Grants

• ERC Proof of Concept grant (co-applicant, 2022)

Publications