ABSTRACT
Simulation Decomposition (SimDec) provides an approach for analyzing computational model behaviour and supporting decision-making. At its core is a visualization of model output decomposed by its most influential input variables, where the noteworthy variables have been identified by the calculation of global sensitivity indices. This chapter explains the overall SimDec algorithm, establishes guidelines for its effective implementation, provides the essential building blocks for understanding and interpreting its results, outlines a novel method for calculating sensitivity indices, describes the open-source packages, and discusses various nuances for implementing effective SimDec studies.
