Categories
Events

VIRTUAL EVENT – FITTING MODELS AND QUANTIFYING THE UNCERTAINTY OF PREDICTIONS WORKSHOP

Wednesday 15th July 2-4 p.m.

This is an addition to our Virtual Event series and is a workshop style event for those interested in:

Fitting Models and Quantifying the Uncertainty of Predictions.

Aim and Scope: The Landscape Decisions Programme contains a number of projects that involve the use of computer models designed to aid the decision-making process. This workshop will bring together researchers working with models, and in particular, will discuss the issues of calibrating models (also known as “parameter estimation/tuning/fitting/optimisation”, or as the “inverse problem”) and comparing or combining models. For this workshop we intend to focus specifically on stochastic models that cannot be fit using standard likelihood-based inference techniques.

This 2-hour zoom event run by Richard Everitt and Richard Sibly will have the following structure: 

  1. Organisers’ welcome and introduction (5 min).
  • Individual case studies. Each will consist of a brief description by a participant of their model, and the challenges associated with estimating its parameters and quantifying the uncertainty attached to its predictions (max 5 min). This will be followed by a discussion of how new inference techniques might help.
  • Conclusion and proposals for future collaborations.

Call for Participation: If you are interested in speaking at this workshop, please by Friday 3rd July fill in the form at https://warwick.ac.uk/fac/sci/statistics/staff/academic-research/everitt/landscape_workshop. If you are interested in parameter estimation, assessing model fit, or uncertainty quantification, but are not sure if your work fits into the scope of this workshop, please get in touch via the form above, and let us know what you are working on. We would like to get an idea of the different types of models being investigated in the Programme and where applicable of the progress being made on parameter estimation and uncertainty quantification. We hope from this to see where future collaborations might be helpful.

Any questions, please email Richard.Everitt@warwick.ac.uk or r.m.sibly@reading.ac.uk.