Managing Uncertainty in Complex Models

This book describes statistical methods for analyzing uncertainty in complex mathematical or computational models of physical systems. Such models are widely used in science and engineering-for example, climate models or cost-effectiveness models in health economics-and almost always have uncertainty in their predictions. Model users may not have all the information they need to specify values of various quantities required by the model, and models usually cannot be expected to represent the real world perfectly. This book presents methods for analyzing uncertainty within the statistical frameworks of Bayesian and Bayes linear analysis. The core of this book is the problem of analyzing uncertainty for computationally expensive models. The book uses simple examples to illustrate the methods and includes substantive case studies to highlight applications.