Project IV (MATH 4072) 2013-14


Bayesian History Matching

Joint Supervisors: Ian Vernon and Michael Goldstein

Description

Many problems in science and technology, involving forecasting and control of physical systems, are addressed by the analysis of computer simulators for the systems. Before we can use the simulator to forecast future outcomes, we often need to tune the simulator input parameters so that the outputs match historical data up to error tolerances. The process of searching for all input parameter choices which give acceptable matches is termed history matching, a method that has been successfully applied across a wide variety of scientific areas including the analysis of climate models, galaxy formation simulations and oil reservoirs. This project is an exploration of the techniques used to carry out this process, which is based on statistical modelling and inversion of the computer simulator.

The first term will be an introduction to the basic methodology involved in history matching, in the context of some simple computer models constructed for this purpose. The approach raises many practical and theoretical questions as to how best to carry out this process, and the second term will concern exploration of various aspects of these questions.

Prerequisites

Statistical Concepts II and Statistical Methods III

Resources

For an introduction to History Matching as applied to a complex model of Galaxy Formation see our paper entitled "Galaxy Formation: Bayesian History Matching for the Observable Universe" which can be found at Statistical Science Future Papers.

An excellent web-site which describes (in sometimes overwhelming detail!) the types of analyses which this project gives an introduction to is:

The MUCM Web-site

This is the web-site for the Managing Uncertainty in Complex Models (MUCM) project, a consortium in which we were involved, (with the Universities of Sheffield, Aston, LSE and Southampton). There are an enormous number of links to follow at this site. One in particular, which gives an introduction to emulation, is:

O'Hagan, A. (2006). Bayesian analysis of computer code outputs: a tutorial. Reliability Engineering and System Safety 91, 1290–1300.

See also the MUCM toolkit for a detailed list of emulation and History Matching related techniques and tools.

email: Ian Vernon


Back