Abstract

Understanding the brain’s complex, high-dimensional processes remains a grand challenge for both neuroscience and computational modeling. While data-driven (bottom-up) approaches provide valuable insights, they are insufficient on their own to deal with the brain’s complexity. In this talk, I will present a complementary “top-down” framework based on normative models — computational theories that specify what the brain should do to perform optimally under given constraints. Focusing on rapid perceptual and value-based decisions, I will demonstrate how such decisions can be formalized as Bayesian decision problems: an agent accumulates noisy evidence through statistical inference and commits to a choice when a decision threshold is reached. I will illustrate these principles using increasingly complex examples and discuss how these models align with observed behavioral and neural data in both humans and animals. Overall, the talk will demonstrate how foundational concepts from artificial intelligence - such as Bayesian decision theory - can offer new insights into the computational principles that govern biological intelligence.

About Jan Drugowitsch

Jan Drugowitsch is Associate Professor of Neurobiology at Harvard Medical School in Boston (US). He leads a research lab that investigates the algorithms and neural mechanisms underlying decision-making and navigation under uncertainty. The lab develops theories by combining tools from machine learning, physics, and our knowledge of the brain’s biological foundations, and tests the arising models and hypotheses in close collaboration with experimental labs. Dr. Drugowitsch draws upon his academic background, including a PhD in probabilistic machine learning from the University of Bath, UK, and extensive postdoctoral training in computational and cognitive neuroscience, to tackle these fundamental inquiries into the operating principles of the nervous system.

About Current Trends in Computer Science

This lecture is part of the Current Trends in Computer Science Lecture Series by the TU Wien Informatics Doctoral School, where renowned Guest Professors hold public lectures every semester. If you are studying with us, the lecture series can be credited as an elective course for students of master programs of computer science: 195.072 Current Trends in Computer Science. Additionally, you can join courses held by this year’s Guest Professors of our doctoral colleges and the TU Wien Informatics Doctoral School.

Curious about our other news? Subscribe to our news feed, calendar, or newsletter, or follow us on social media.