ANOVA decomposition of conditional Gaussian processes for sensitivity analysis with dependent inputs

TL;DR

This paper introduces a novel ANOVA decomposition for Gaussian processes to efficiently perform sensitivity analysis with dependent inputs, reducing computational costs for complex models.

Contribution

It proposes a new high-dimensional Gaussian process decomposition and sensitivity indices tailored for dependent inputs, with estimation methodology and uncertainty quantification.

Findings

Effective sensitivity indices for dependent inputs

Validated on toy functions and flood model

Reduces number of runs needed for analysis

Abstract

Complex computer codes are widely used in science to model physical systems. Sensitivity analysis aims to measure the contributions of the inputs on the code output variability. An efficient tool to perform such analysis are the variance-based methods which have been recently investigated in the framework of dependent inputs. One of their issue is that they require a large number of runs for the complex simulators. To handle it, a Gaussian process regression model may be used to approximate the complex code. In this work, we propose to decompose a Gaussian process into a high dimensional representation. This leads to the definition of a variance-based sensitivity measure well tailored for non-independent inputs. We give a methodology to estimate these indices and to quantify their uncertainty. Finally, the approach is illustrated on toy functions and on a river flood model.