Events Calendar

Generative Models and the Transport of Measure

Speaker: Gavin Kerrigan, PhD Candidate - Department of Computer Science, UC Irvine

Talk Title: Generative Models and the Transport of Measure

Abstract: A key theme in contemporary generative modeling is the continuous transport of measure, in which a simple reference distribution is gradually transformed into the data distribution. Many recent models, including diffusions and flows, can be viewed through this unifying lens. In this talk, we will first explore some geometric tools for studying dynamics in the space of probability measures. We will then leverage these tools to design generative models, with a focus on applications to inverse problems and complex data structures such as function-valued data.    
 
This lecture satisfies requirements for CSCI 591: Research Colloquium.
 
In-person ONLY; recording available post-presentation.
 
 

Biography: Gavin Kerrigan is a final year PhD candidate in the Department of Computer Science at UC Irvine, where he is advised by Padhraic Smyth. Prior to joining UCI, he obtained a BSc in mathematics from the Schreyer Honors College at Penn State University. His research focuses on advancing the theory and practice of deep generative models, ranging from fundamental methodology to applications in climate science. His work has been recognized through a best paper award at AISTATS'23 for contributions to function-space generative modeling.

Host: USC Machine Learning Center

Location: Ronald Tutor Hall of Engineering (RTH) - 211

Audiences: Everyone Is Invited

Contact: Thomas Lord Department of Computer Science