A Review of Global Sensitivity Analysis Methods and a comparative case study on Digit Classification

TL;DR

This study goes through the underlying mechanism of widely used GSA methods and highlights their efficacy through a comprehensive methodology and proposes a methodology for evaluating the efficacy of these methods by conducting a case study on MNIST digit dataset.

Abstract

Global sensitivity analysis (GSA) aims to detect influential input factors that lead a model to arrive at a certain decision and is a significant approach for mitigating the computational burden of processing high dimensional data. In this paper, we provide a comprehensive review and a comparison on global sensitivity analysis methods. Additionally, we propose a methodology for evaluating the efficacy of these methods by conducting a case study on MNIST digit dataset. Our study goes through the underlying mechanism of widely used GSA methods and highlights their efficacy through a comprehensive methodology.

Figures (4)

• Figure 1: Methodology for comparing SA methods
• Figure 2: Accuracy of MNIST classification. a: b:
• Figure 3: Comparison between the most influential and non-influential factors
• Figure 4: Visualization of the sensitivity indices of each method trained on MNIST classification task