-
Gaussian linear approximation for the estimation of the Shapley effects
Authors:
Baptiste Broto,
François Bachoc,
Marine Depecker,
Jean-Marc Martinez
Abstract:
In this paper, we address the estimation of the sensitivity indices called "Shapley eects". These sensitivity indices enable to handle dependent input variables. The Shapley eects are generally dicult to estimate, but they are easily computable in the Gaussian linear framework. The aim of this work is to use the values of the Shapley eects in an approximated Gaussian linear framework as estimators…
▽ More
In this paper, we address the estimation of the sensitivity indices called "Shapley eects". These sensitivity indices enable to handle dependent input variables. The Shapley eects are generally dicult to estimate, but they are easily computable in the Gaussian linear framework. The aim of this work is to use the values of the Shapley eects in an approximated Gaussian linear framework as estimators of the true Shapley eects corresponding to a non-linear model. First, we assume that the input variables are Gaussian with small variances. We provide rates of convergence of the estimated Shapley eects to the true Shapley eects. Then, we focus on the case where the inputs are given by an non-Gaussian empirical mean. We prove that, under some mild assumptions, when the number of terms in the empirical mean increases, the dierence between the true Shapley eects and the estimated Shapley eects given by the Gaussian linear approximation converges to 0. Our theoretical results are supported by numerical studies, showing that the Gaussian linear approximation is accurate and enables to decrease the computational time signicantly.
△ Less
Submitted 3 June, 2020;
originally announced June 2020.
-
Variance reduction for estimation of Shapley effects and adaptation to unknown input distribution
Authors:
Baptiste Broto,
François Bachoc,
Marine Depecker
Abstract:
The Shapley effects are global sensitivity indices: they quantify the impact of each input variable on the output variable in a model. In this work, we suggest new estimators of these sensitivity indices. When the input distribution is known, we investigate the already existing estimator and suggest a new one with a lower variance. Then, when the distribution of the inputs is unknown, we extend th…
▽ More
The Shapley effects are global sensitivity indices: they quantify the impact of each input variable on the output variable in a model. In this work, we suggest new estimators of these sensitivity indices. When the input distribution is known, we investigate the already existing estimator and suggest a new one with a lower variance. Then, when the distribution of the inputs is unknown, we extend these estimators. Finally, we provide asymptotic properties of the estimators studied in this article.
△ Less
Submitted 13 February, 2020; v1 submitted 21 December, 2018;
originally announced December 2018.
-
Sensitivity indices for independent groups of variables
Authors:
Baptiste Broto,
François Bachoc,
Marine Depecker,
Jean-Marc Martinez
Abstract:
In this paper, we study sensitivity indices for independent groups of variables and we look at the particular case of block-additive models. We show in this case that most of the Sobol indices are equal to zero and that Shapley effects can be estimated more efficiently. We then apply this study to Gaussian linear models, and we provide an efficient algorithm to compute the theoretical sensitivity…
▽ More
In this paper, we study sensitivity indices for independent groups of variables and we look at the particular case of block-additive models. We show in this case that most of the Sobol indices are equal to zero and that Shapley effects can be estimated more efficiently. We then apply this study to Gaussian linear models, and we provide an efficient algorithm to compute the theoretical sensitivity indices. In numerical experiments, we show that this algorithm compares favourably to other existing methods. We also use the theoretical results to improve the estimation of the Shapley effects for general models, when the inputs form independent groups of variables.
△ Less
Submitted 11 December, 2018; v1 submitted 12 January, 2018;
originally announced January 2018.
