Sensitivity Analysis of Image Classification Models using Generalized Polynomial Chaos

Lukas Bahr, Lucas Poßner, Konstantin Weise, Sophie Gröger, Rüdiger Daub
6/23/2025
cs.LGcs.AI

No official code found. Best implementation: lpossner/gPC-Image-Classification (1 star, MIT, last commit 2026-01-01).

Abstract

Integrating advanced communication protocols in production has accelerated the adoption of data-driven predictive quality methods, notably machine learning (ML) models. However, ML models in image classification often face significant uncertainties arising from model, data, and domain shifts. These uncertainties lead to overconfidence in the classification model's output. To better understand these models, sensitivity analysis can help to analyze the relative influence of input parameters on the output. This work investigates the sensitivity of image classification models used for predictive quality. We propose modeling the distributional domain shifts of inputs with random variables and quantifying their impact on the model's outputs using Sobol indices computed via generalized polynomial chaos (GPC). This approach is validated through a case study involving a welding defect classification problem, utilizing a fine-tuned ResNet18 model and an emblem classification model used in BMW Group production facilities.

View on arXivView PDF

Code Implementations(1)

lpossner/gPC-Image-ClassificationCommunity72%
MIT

This repository contains the code for the research paper titled "Sensitivity Analysis of Image Classification Models using Generalized Polynomial Chaos", which is currently published as a preprint on arXiv. Here, we provide all the necessary information and resources to reproduce the results presented in our paper.

10PythonJan 24, 20259 months agoMIT

Similar Papers

Model Monitoring and Robustness of In-Use Machine Learning Models: Quantifying Data Distribution Shifts Using Population Stability Index

79% match2023

Self-Ensemble Post Learning for Noisy Domain Generalization

79% match2025

Improving Supervised Machine Learning Performance in Optical Quality Control via Generative AI for Dataset Expansion

79% match2026

Predicting the Reliability of an Image Classifier under Image Distortion

79% match2024

Enhancing Variable Importance in Random Forests: A Novel Application of Global Sensitivity Analysis

78% match2024

Explaining an image classifier with a generative model conditioned by uncertainty

78% match2024

Cite this paper

@article{bahr2025sensitivity,
  title  = {Sensitivity Analysis of Image Classification Models using Generalized Polynomial Chaos},
  author = {Lukas Bahr and Lucas Poßner and Konstantin Weise and Sophie Gröger and Rüdiger Daub},
  year   = {2025},
  eprint = {2506.18751},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2506.18751v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2506.18751.svg)](https://codeofpaper.com/paper/2506.18751)

Discussion