Webinar on Introduction to Global Sensitivity Analysis and Metamodelling
7th February 2019 at 3pm (UK time)
High complexity of models in physics, chemistry, bioengineering, environmental studies and other fields results in the increased uncertainty in model parameters, model structures and model predictions. Sensitivity analysis (SA) aims at quantifying the relative importance of input parameters in determining values of model outputs. Global SA (GSA) estimates the effect of varying a given input (or set of inputs) while all other inputs are varied as well, thus providing a measure of interactions among them. GSA is used to identify key inputs whose uncertainty most affects the output and the results are used to rank inputs, fix unessential inputs and decrease problem dimensionality. Over the last decade GSA has gained acceptance among practitioners in the process of model development, calibration and validation, reliability and robustness analysis, decision-making under uncertainty, quality-assurance, and complexity reduction. In this presentation, we will provide an overview of methods and techniques of GSA.
For computationally expensive models and models which need to be run repeatedly on-line the replacement of complex models by equivalent operational metamodels (also known as surrogate models) is a practical way of making computations tractable. Such an approach is also invaluable in cases when the model is not given explicitly but it exists only in a form of input- output maps or is given as a “black box”. Another advantage of using metamodeling methods is that they can provide the values of GSA measures without the need of extra model runs. In the second part of this presentation, we will discuss some of the metamodelling methods and software tolls that we have been developing at Imperial College London over the last decade.
This webinar will be presented by Sergei Kucherenko of Imperial College London and should last no longer than one hour.
If you wish to register please contact Christine Stevenson at Christine.email@example.com