BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Physics-aware data-driven approaches for time prediction - Luca Ma
 gri\, Imperial College London
DTSTART:20230609T150000Z
DTEND:20230609T160000Z
UID:TALK200761@talks.cam.ac.uk
CONTACT:Prof. Jerome Neufeld
DESCRIPTION:To predict the evolution of physical systems\, we need a model
  that tells us “what happens next” given “what we know so far”. Th
 is can be enabled by physical principles and data-driven approaches. \n \n
 On the one hand\, physical principles\, for example conservation laws\, ar
 e extrapolative because they can provide predictions on phenomena that hav
 e not been observed\, but they are “rigid”. On the other hand\, data-d
 riven modelling provides correlation functions within data\, but they are 
 “adaptive”. In this talk\, the complementary capabilities of both appr
 oaches will be exploited to achieve adaptive modelling and optimization of
  nonlinear\, unsteady\, and uncertain flows. \n \nThe focus of the talk is
  on computational methodologies for modelling and optimization of complex 
 flows: \n(i) and auto-encoders and reservoir computers for reduced-order m
 odelling of turbulent flows\, which generalise POD/DMD methods to nonlinea
 r dynamics\, for the prediction of extreme events\; and\n(ii) real-time da
 ta assimilation with a Bayesian approach to infer model errors (bias) with
  applications to thermoacoustic oscillations (time permitting).
LOCATION:MR2
END:VEVENT
END:VCALENDAR