Home // DATA ANALYTICS 2022, The Eleventh International Conference on Data Analytics // View article

Shapley Values based Regional Feature Importance Measures Driving Error Analysis in Manufacturing

Authors:
Valentin Göttisheim
Holger Ziekow
Ulf Schreier
Alexander Gerling

Keywords: manufacturing quality management; error analysis; feature importance; Shapley Values; xAI; machine learning.

Abstract:
Data driven manufacturing quality management using machine learning for error detection can leverage predictive models for error analysis. Quality engineer experts evaluate the models input and interpret important features in the context of the specific manufacturing domain. In this paper, we propose three heuristics to determine the importance of features leading to actionable insights for error analysis. All proposed metrics are illustrated on synthetic data and evaluated on a real-world dataset.

Pages: 19 to 26

Copyright: Copyright (c) IARIA, 2022

Publication date: November 13, 2022

Published in: conference

ISSN: 2308-4464

ISBN: 978-1-61208-994-2

Location: Valencia, Spain

Dates: from November 13, 2022 to November 17, 2022