BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Calibrating Data-Driven Predictions for Safety-Critical Systems: C
 hallenges and Solutions - Carla Ferreira (Durham University)
DTSTART:20250603T112500Z
DTEND:20250603T114500Z
UID:TALK230803@talks.cam.ac.uk
DESCRIPTION:As safety-critical systems&mdash\;ranging from autonomous tran
 sport to industrial control systems&mdash\;become increasingly data-driven
 \, ensuring reliable probabilistic predictions is a fundamental challenge.
  Historically\, the safety of such systems has been ensured through physic
 s-based modeling\, scenario analysis\, and conservative engineering design
 . However\, as machine learning (ML) models are increasingly used for pred
 ictive decision-making\, they introduce additional uncertainties that must
  be well-calibrated to maintain system reliability.\nThis talk explores th
 e role of probabilistic calibration techniques in improving the trustworth
 iness of ML-based predictions in safety-critical applications. We will dis
 cuss:\n\n\nThe challenges of uncalibrated ML models in high-risk environme
 nts.\n\n\nApproaches for calibrating ML predictions\, from conformal predi
 ction to Bayesian calibration.\n\n\nThe role of uncertainty-aware experime
 ntal design to reduce uncertainty in safety-critical applications.\n\n\nHy
 brid approaches that combine physics-based models with data-driven insight
 s to ensure robustness.\n\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR