Events
FDS Postdoctoral Applicants
Large-scale probabilistic inference via optimal transport
 |
Speaker: Aram-Alexandre Pooladian (NYU) New York University Wednesday, December 18, 2024 1:00PM - 2:00PM and via Webcast: https://yale.zoom.us/j/97644284302 |
“Large-scale probabilistic inference via optimal transport”
Abstract: Many contemporary challenges in probabilistic inference boil down to understanding how measures evolve. For generative models or in trajectory inference, we ask how samples, representing distributions, transform over time. When optimizing a functional over the space of measures, the optimization iterations themselves concern the transformations of measures. In this talk, I will use optimal transport (OT) as a unifying framework to address these challenges in data science. Part I is about optimization guarantees via OT, and Part II surrounds statistical guarantees for generative modeling. This is joint work with Roger Jiang (NYU), Sinho Chewi (IAS –> Yale), and Jonathan Niles-Weed (NYU).
Bio: Aram-Alexandre Pooladian is a PhD Candidate at New York University where he is supervised by Jonathan Niles-Weed. His research incorporates optimal transport as a tool for understanding the mathematics of data science and for large-scale applications in machine learning. His research is funded by the National Science Foundation, the National Science and Engineering Council of Canada, Google, and Meta AI. His research has received a Best Paper Award from NeurIPS — Optimal Transport and Machine Learning.
Add To: Google Calendar | Outlook | iCal File
Submit an Event
Interested in creating your own event, or have an event to share? Please fill the form if you’d like to send us an event you’d like to have added to the calendar.