Please welcome Dr. Matthew Plumlee from Northwestern University. His topic of discussion is inexact computer model calibration: Concerns, controversy, credibility, and confidence.
As always refreshments are available in 428 Daniels Hall 30 minutes before the seminar begins.
There has been a recent surge in statistical methods for calibration of inexact models. An inexact model is one
that disagrees with the true system for some subset of inputs. Examples include models of weather systems, finite element models of cars, or chemical reaction models. Alongside these developments, a controversy has emerged about the goals of calibration of inexact models. This talk will trace a swath of research stemming from twenty years ago and potential concerns are marked along the way. The most notable problem is identifiability: if you have unknown parameters in your model alongside an unknown model discrepancy, it seems impossible to learn both quantities. The talk will also present some new ideas in this setting that might help close some of these philosophical and practical issues.
Matthew Plumlee joined Northwestern IEMS in the Fall of 2017. His interests are in the interface of data and modeling; specifically in methods for experimentation and uncertainty quantification for complex systems. This includes: model calibration; design and analysis of computer experiments; large-scale simulation and experimentation; stochastic modeling of enterprise, mechanical and biological systems; and general analytical/statistical methods and computational techniques. He holds a BA in Mechanical Engineering from Purdue University, and an MS and a PhD from the Georgia Institute of Technology in Statistics and Industrial Engineering, respectively