A 'Propagative' Approach to Sensitivity Analysis.
Abstract
Large-scale simulations often involve huge numbers of parameters, making it prohibitive to run more than a tiny fraction of all potentially relevant cases. Sensitivity analysis attempts to show how sensitive the results of a simulation are to changes in it parameters; this is an important tool for promoting confidence in a simulation and making its results credible. However, the computational cost of traditional approaches to sensitivity analysis prevents its use in many cases. We show that this cost is logically unnecessary and can be largely avoided by propagating and combining sensitivities during a computation, rather than recomputing them. We describe this propagative approach to sensitivity analysis and present the algorithm we have implemented to explore its potential.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1990
- Accession Number
- ADA239933
Entities
People
- Charlene A. Hefley
- Jeff Rothenberg
- Norman Z. Shipiro
Organizations
- RAND Corporation