Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability

Understanding and interpreting machine learning models is crucial in high-stakes industries like steel manufacturing, where decisions impact energy efficiency and environmental sustainability. Traditional feature importance methods often struggle with robustness under noisy conditions, leading to un...

Full description

Saved in:
Bibliographic Details
Main Authors: Grigorios Tzionis, Georgia Kougka, Ilias Gialampoukidis, Stefanos Vrochidis, Ioannis Kompatsiaris, Maro Vlachopoulou
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Applied Artificial Intelligence
Online Access:https://www.tandfonline.com/doi/10.1080/08839514.2025.2515062
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849727488483131392
author Grigorios Tzionis
Georgia Kougka
Ilias Gialampoukidis
Stefanos Vrochidis
Ioannis Kompatsiaris
Maro Vlachopoulou
author_facet Grigorios Tzionis
Georgia Kougka
Ilias Gialampoukidis
Stefanos Vrochidis
Ioannis Kompatsiaris
Maro Vlachopoulou
author_sort Grigorios Tzionis
collection DOAJ
description Understanding and interpreting machine learning models is crucial in high-stakes industries like steel manufacturing, where decisions impact energy efficiency and environmental sustainability. Traditional feature importance methods often struggle with robustness under noisy conditions, leading to unreliable insights. To address this problem, we introduce the Complexity and Interaction Enhanced SHAP (CESHAP), a novel feature importance method that incorporates model complexity and feature interactions. Alongside, we propose the Noise-Adjusted Feature Importance Change (NAFIC) metric to assess the robustness of feature importance methods against varying levels of noise. Experiments conducted on an energy consumption dataset from the steel industry, with systematically introduced Gaussian noise levels (5%, 10%, 15%, 20%), demonstrate that CESHAP offers superior robustness in tree-based models compared to SHapley Additive exPlanations (SHAP) and Permutation Feature Importance (PFI). Our findings underscore the effectiveness of CESHAP in enhancing interpretability and reliability in complex, non-linear models, ultimately supporting more informed decision-making in energy-intensive industries.
format Article
id doaj-art-bef0097dc54f4ddeae08ca14000f5915
institution DOAJ
issn 0883-9514
1087-6545
language English
publishDate 2025-12-01
publisher Taylor & Francis Group
record_format Article
series Applied Artificial Intelligence
spelling doaj-art-bef0097dc54f4ddeae08ca14000f59152025-08-20T03:09:49ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452025-12-0139110.1080/08839514.2025.2515062Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved InterpretabilityGrigorios Tzionis0Georgia Kougka1Ilias Gialampoukidis2Stefanos Vrochidis3Ioannis Kompatsiaris4Maro Vlachopoulou5Information Technologies Institute (ITI), Centre For Research and Technology Hellas (CERTH), Thessaloniki, GreeceInformation Technologies Institute (ITI), Centre For Research and Technology Hellas (CERTH), Thessaloniki, GreeceInformation Technologies Institute (ITI), Centre For Research and Technology Hellas (CERTH), Thessaloniki, GreeceInformation Technologies Institute (ITI), Centre For Research and Technology Hellas (CERTH), Thessaloniki, GreeceInformation Technologies Institute (ITI), Centre For Research and Technology Hellas (CERTH), Thessaloniki, GreeceDepartment of Applied Informatics, University of Macedonia, Thessaloniki, GreeceUnderstanding and interpreting machine learning models is crucial in high-stakes industries like steel manufacturing, where decisions impact energy efficiency and environmental sustainability. Traditional feature importance methods often struggle with robustness under noisy conditions, leading to unreliable insights. To address this problem, we introduce the Complexity and Interaction Enhanced SHAP (CESHAP), a novel feature importance method that incorporates model complexity and feature interactions. Alongside, we propose the Noise-Adjusted Feature Importance Change (NAFIC) metric to assess the robustness of feature importance methods against varying levels of noise. Experiments conducted on an energy consumption dataset from the steel industry, with systematically introduced Gaussian noise levels (5%, 10%, 15%, 20%), demonstrate that CESHAP offers superior robustness in tree-based models compared to SHapley Additive exPlanations (SHAP) and Permutation Feature Importance (PFI). Our findings underscore the effectiveness of CESHAP in enhancing interpretability and reliability in complex, non-linear models, ultimately supporting more informed decision-making in energy-intensive industries.https://www.tandfonline.com/doi/10.1080/08839514.2025.2515062
spellingShingle Grigorios Tzionis
Georgia Kougka
Ilias Gialampoukidis
Stefanos Vrochidis
Ioannis Kompatsiaris
Maro Vlachopoulou
Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
Applied Artificial Intelligence
title Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
title_full Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
title_fullStr Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
title_full_unstemmed Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
title_short Enhancing Robustness in Feature Importance Methods with NAFIC and CESHAP for Improved Interpretability
title_sort enhancing robustness in feature importance methods with nafic and ceshap for improved interpretability
url https://www.tandfonline.com/doi/10.1080/08839514.2025.2515062
work_keys_str_mv AT grigoriostzionis enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability
AT georgiakougka enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability
AT iliasgialampoukidis enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability
AT stefanosvrochidis enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability
AT ioanniskompatsiaris enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability
AT marovlachopoulou enhancingrobustnessinfeatureimportancemethodswithnaficandceshapforimprovedinterpretability