BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Testing and improving the robustness of amortized Bayesian inferen
 ce for cognitive models - Yufei Wu (KU Leuven)
DTSTART:20250625T091500Z
DTEND:20250625T094500Z
UID:TALK232306@talks.cam.ac.uk
DESCRIPTION:Author: Yufei Wu\, Stefan Radev\, Francis Tuerlinckx\nThis pap
 er addresses the challenge of achieving robust and reliable parameter esti
 mation in cognitive models when contaminations and outliers are present. S
 tudies have demonstrated that amortized Bayesian inference (ABI) has the p
 otential to facilitate inference of arbitrarily complex cognitive models v
 ia simulations and neural networks. However\, it remains unclear whether A
 BI is robust to outliers and how it can be robustified. Our findings show 
 that by simulating a more realistic data-generation process that incorpora
 tes contaminants and using the contaminated data to train the neural netwo
 rks\, the inference with ABI achieves a high degree of robustness while ma
 intaining efficiency. These results are significant because they provide a
  powerful yet practical solution to handle outliers in empirical research 
 with ABI.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR