BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260412T235106EDT-0189WRTG9D@132.216.98.100
DTSTAMP:20260413T035106Z
DESCRIPTION:Title: Artificially Intelligent Geospatial Systems: A Case Stud
 y in Energetics for Mobile Health Data.\n\nAbstract: I will share my persp
 ectives on the significant paradigm shift taking place in data analysis wi
 th the advent of AI technologies. This rapidly evolving field offers subst
 antial intellectual space for statistical theory and methods to not only c
 o-exist with other disciplines within computer science and machine learnin
 g\, but also play a crucial role in advancing data analysis and probabilis
 tic inference at unprecedented scales. I will elucidate three ideas that w
 ill synthesize into an artificially intelligent inferential system. The fi
 rst is 'amortized Bayesian inference' that considers training and calculat
 ing posterior distributions using generative AI. The second is Bayesian tr
 ansfer learning for scaling Inference to massive datasets. The third is Ba
 yesian predictive stacking that delivers exact simulation-based inference 
 without resorting to expensive iterative methods such as Markov chain Mont
 e Carlo. I will base my talk on a case study that is a part of the Univers
 ity of California Los Angeles (UCLA) Physical Activity and Sustainable Tra
 nsportation Approaches (PASTA-LA) and is primarily concerned with learning
  about a subject's metabolic levels as a function of their mobility attrib
 utes and other health attributes.\n
DTSTART:20251121T203000Z
DTEND:20251121T213000Z
LOCATION:Room 1140\, McGill College 2001\, CA\, QC\, Montreal\, H3A 1G1\, 2
 001\, avenue McGill College
SUMMARY:Sudipto Banerjee (UCLA)
URL:https://www.mcgill.ca/mathstat/channels/event/sudipto-banerjee-ucla-368
 803
END:VEVENT
END:VCALENDAR