Random balance designs for the estimation of first order global sensitivity indices

Stefano Tarantola (1) , Debora Gatelli (2) , Thierry A. Mara (3)

1 Institute for the Protection and Security of the Citizen
2 Institue for protection and Security of the Citizen
3 PIMENT - Physique et Ingénierie Mathématique pour l'Énergie, l'environnemeNt et le bâtimenT

We present two methods for the estimation of main effects in global sensitivity analysis. The methods adopt Satterthwaite's application of random balance designs in regression problems, and extend it to sensitivity analysis of model output for non-linear, non-additive models. Finite as well as infinite ranges for model input factors are allowed. The methods are easier to implement than any other method available for global sensitivity analysis, and reduce significantly the computational cost of the analysis. We test their performance on different test cases, including an international benchmark on safety assessment for nuclear waste disposal originally carried out by OECD/NEA.

Mots clés: Computational models; Uncertainty analysis; Global sensitivity analysis; Fourier amplitude sensitivity test

Domaines: Applications [stat.AP]

Stefano Tarantola, Debora Gatelli, Thierry A. Mara. Random balance designs for the estimation of first order global sensitivity indices. Reliability Engineering and System Safety, 2006, 91 (6), pp.717-727. ⟨10.1016/j.ress.2005.06.003⟩. ⟨hal-01065897⟩