Virtual Talk with Julian Gerstenberg
Dr. Julian Gerstenberg, Interim Professor of Stochastics at Philipps University Marburg.
Probability Distributions in Machine Learning: Beyond Expected Values and Across Domains
The concept of a probability distribution is foundational in numerous scientific disciplines. This presentation provides an accessible overview of their crucial role in context of computational methods, in particular in different subareas of machine learning. Touching on topics like Distributional Reinforcement Learning and the parallels between nonparametric statistics and data-driven learning, it also delves into challenges with relational data, such as graph data, e.g. appearing in social network analysis. Throughout, theoretical insights are coupled with practical experiences, emphasizing the nuanced role of probability distributions in machine learning applications.
Julian Gerstenberg serves as an interim professor of stochastics at Philipps University Marburg and leads the DFG-funded project "Exchangeability Theory of ID-Based Data Structures with Applications in Statistics". His research focuses on the intersections of nonparametric statistics, machine learning, and data structures, combining rigorous mathematical theories with applied topics such as software design and validation protocols.
After receiving his PhD in probability theory in 2018 from Leibniz University Hanover, he joined the AI consulting and engineering company AMAI as a data scientist. There, he gained hands-on expertise in the machine learning workflow, which continues to influence his research interests. As an educator, he has taught diverse courses and mentored students, emphasising and developing the stochastic foundations central to modern machine learning techniques.
Please register with firstname.lastname@example.org to receive the Zoom login.