Sabina Sloman

Sabina Sloman

she/her

Assistant Professor in Statistics & Data Science @ University of Birmingham

About me

I am an assistant professor in the Statistics and Data Science Group at the University of Birmingham and a member of the ELLIS Society.

I received my PhD in Social and Decision Sciences (concentration in Cognitive Decision Sciences) in 2022 from Carnegie Mellon University, where I was advised by Dr. Daniel Oppenheimer. My dissertation investigated the robustness of Bayesian experimental design to misspecification. From 2023 to 2026, I was a postdoc with the Manchester Centre for AI Fundamentals at the University of Manchester, supervised by Dr. Samuel Kaski.

During my PhD, I worked on methods for robust cognitive modeling. I have since extended this work to better understand and address the challenges to robust statistical learning more broadly. Most of my work uses the framework of Bayesian inference. I am particularly interested in the following research areas: model misspecification; experimental design; parameter identifiability and correlation; robustness; transfer learning; statistical learning theory; uncertainty quantification. I maintain a strong interest in the applications of my work to scientific theory and practice, particularly in cognitive science.

I am grateful to be supported, inspired, and taught by my collaborators, including members of the Finnish Center for Artificial Intelligence, Cognitive Epistemology Lab at UC Berkeley, and Global Risk and Individual Decisions Laboratory at Purdue University; and PhD students Yasir Barlas and Roubing Tang.

Research

Statistical learning is the process of resolving uncertainty about which of a set of candidate models best corresponds to a target system, or data-generating process. It is the foundation of scientific discovery, machine learning, and human cognition. If the investigator’s model class accounts for all sources of uncertainty, existing statistical learning frameworks are remarkably powerful in helping scientists, algorithms, and humans better understand and navigate the world.

My work takes a data-centric perspective on learning in the sense that I focus on (i) how the structure of the available data affects whether and how much learning occurs, and (ii) the development of methods that alter the structure of one’s data in ways that promote learning (e.g., active learning methods).

My recent research is driven by questions like:

  • How can active learning methods effectively balance the resolution of multiple forms of uncertainty (e.g., between the values of target and nuisance parameters)?
  • What are the causes and consequences of negative transfer in applying knowledge gleaned in particular source environments to a similar-but-distinct target environment?
  • When and how does increasing the expressivity of a (cognitive or artificial) system improve its ability to learn?

Contact

I am always enthusiastic to discuss research or answer questions! You can reach me at s (dot) lastname (at) bham (dot) ac (dot) uk.

News