Uncertainty impacts decision makers across government and industry. However, calculation and clear communication of uncertainty poses significant challenges, as highlighted by Prof. Sir Mark Walport in a recent article1.
This conference will bring together decision makers, analysts and researchers from government, industry and academia to explore state of the art techniques and common problems, and share best practice.
The event will cover the methods for calculating and propagating uncertainty within complex systems of mathematical or computer models and simulations. It will also explore the communication of uncertainty and risk to decision makers.
Sources of uncertainty are seen to include human input, human elicited input, intrinsic and irreducible uncertainty in model inputs, calibration and tuning errors, model error, lack of training data and ill-defined or unknown problem spaces, each of these sources need to be explained and accounted for in order to provide reliable advice. The careful communication of these uncertainties to non-specialists then becomes critical and this will also be explored in detail.
The first day will be structured as a conference, with an introduction by Prof. Bernard Silverman (Chief Scientific Advisor to the Home Office) and a keynote lecture by Prof. Sir David Spiegelhalter.
The second day will be structured as four parallel workshops.
Abstract submission: 1 October 2014
Registration: 1 November 2014
The number of participants is limited and so early registration is strongly advised.
The outline programme for the conference is now available.
For further information, please contact the conference organisers.1 Communicating risk and uncertainty in science, FST Journal, May 2014, Vol 21(3).