2016arXiv (Cornell University)Open access

Derivative-based global sensitivity measures and their link with Sobol sensitivity indices

Sergei Kucherenko, Shufang Song

Open full text 5 citations

Abstract

The variance-based method of Sobol sensitivity indices is very popular among practitioners due to its efficiency and easiness of interpretation. However, for high-dimensional models the direct application of this method can be very time consuming and prohibitively expensive to use. One of the alternative global sensitivity analysis methods known as the method of derivative based global sensitivity measures (DGSM) has recently become popular among practitioners. It has a link with the Morris screening method and Sobol sensitivity indices. DGSM are very easy to implement and evaluate numerically. The computational time required for numerical evaluation of DGSM is generally much lower than that for estimation of Sobol sensitivity indices. We present a survey of recent advances in DGSM and new results concerning new lower and upper bounds on the values of Sobol total sensitivity indices. Using these bounds it is possible in most cases to get a good practical estimation of the values of Sobol total sensitivity indices. Several examples are used to illustrate an application of DGSM.

About this research paper

What this paper is about

The variance-based method of Sobol sensitivity indices is very popular among practitioners due to its efficiency and easiness of interpretation. However, for high-dimensional models the direct application of this method can be very time consuming and prohibitively expensive to use. One of the alternative global sensitivity analysis methods known as the method of derivative based global sensitivity measures (DGSM) has recently become popular among practitioners. It has a link with the Morris screening method and Sobol sensitivity indices. DGSM are very easy to implement and evaluate numerically. The computational time required for numerical evaluation of DGSM is generally much lower than that for estimation of Sobol sensitivity indices. We present a survey of recent advances in DGSM and new results concerning new lower and upper bounds on the values of Sobol total sensitivity indices. Using these bounds it is possible in most cases to get a good practical estimation of the values of Sobol total sensitivity indices. Several examples are used to illustrate an application of DGSM.

Why it matters

OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The variance-based method of Sobol sensitivity indices is very popular among practitioners due to its efficiency and easiness of interpretation. However, for high-dimensional models the direct application of this method can be very time consuming and prohibitively expensive to use. One of the alternative global sensitivity analysis methods known as the method of derivative based global sensitivity measures (DGSM) has recently become popular among practitioners. It has a link with the Morris screening method and Sobol sensitivity indices. DGSM are very easy to implement and evaluate numerically. The computational time required for numerical evaluation of DGSM is generally much lower than that for estimation of Sobol sensitivity indices. We present a survey of recent advances in DGSM and new results concerning new lower and upper bounds on the values of Sobol total sensitivity indices. Using these bounds it is possible in most cases to get a good practical estimation of the values of Sobol total sensitivity indices. Several examples are used to illustrate an application of DGSM.

Key concepts: Sobol sequence, Sensitivity (control systems), Variance (accounting), Mathematics, Computer science, Mathematical optimization, Engineering, Economics

Related papers

Back to paper searchBrowse research topicsOriginal source
Derivative-based global sensitivity measures and their link with Sobol sensitivity indices — Research Paper | ScholarLens