On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis

Abstract Machine learning (ML) is increasingly considered the solution to environmental problems where limited or no physico‐chemical process understanding exists. But in supporting high‐stakes decisions, where the ability to explain possible solutions is key to their acceptability and legitimacy, M...

Full description

Saved in:
Bibliographic Details
Main Authors: Banamali Panigrahi, Saman Razavi, Lorne E. Doig, Blanchard Cordell, Hoshin V. Gupta, Karsten Liber
Format: Article
Language:English
Published: Wiley 2025-03-01
Series:Water Resources Research
Subjects:
Online Access:https://doi.org/10.1029/2024WR037398
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849422693701517312
author Banamali Panigrahi
Saman Razavi
Lorne E. Doig
Blanchard Cordell
Hoshin V. Gupta
Karsten Liber
author_facet Banamali Panigrahi
Saman Razavi
Lorne E. Doig
Blanchard Cordell
Hoshin V. Gupta
Karsten Liber
author_sort Banamali Panigrahi
collection DOAJ
description Abstract Machine learning (ML) is increasingly considered the solution to environmental problems where limited or no physico‐chemical process understanding exists. But in supporting high‐stakes decisions, where the ability to explain possible solutions is key to their acceptability and legitimacy, ML can fall short. Here, we develop a method, rooted in formal sensitivity analysis, to uncover the primary drivers behind ML predictions. Unlike many methods for explainable artificial intelligence (XAI), this method (a) accounts for complex multi‐variate distributional properties of data, common in environmental systems, (b) offers a global assessment of the input‐output response surface formed by ML, rather than focusing solely on local regions around existing data points, and (c) is scalable and data‐size independent, ensuring computational efficiency with large data sets. We apply this method to a suite of ML models predicting various water quality variables in a pilot‐scale experimental pit lake. A critical finding is that subtle alterations in the design of some ML models (such as variations in random seed, functional class, hyperparameters, or data splitting) can lead to different interpretations of how outputs depend on inputs. Further, models from different ML families (decision trees, connectionists, or kernels) may focus on different aspects of the information provided by data, despite displaying similar predictive power. Overall, our results underscore the need to assess the explanatory robustness of ML models and advocate for using model ensembles to gain deeper insights into system drivers and improve prediction reliability.
format Article
id doaj-art-7fd72afd5aa94499a7853b07acf751d9
institution Kabale University
issn 0043-1397
1944-7973
language English
publishDate 2025-03-01
publisher Wiley
record_format Article
series Water Resources Research
spelling doaj-art-7fd72afd5aa94499a7853b07acf751d92025-08-20T03:30:57ZengWileyWater Resources Research0043-13971944-79732025-03-01613n/an/a10.1029/2024WR037398On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity AnalysisBanamali Panigrahi0Saman Razavi1Lorne E. Doig2Blanchard Cordell3Hoshin V. Gupta4Karsten Liber5Toxicology Centre University of Saskatchewan Saskatoon SK CanadaSchool of Environment and Sustainability University of Saskatchewan Saskatoon SK CanadaToxicology Centre University of Saskatchewan Saskatoon SK CanadaGlobal Institute for Water Security School of Environmental and Sustainability University of Saskatchewan Saskatoon SK CanadaDepartment of Hydrology and Atmospheric Sciences The University of Arizona Tucson AZ USAToxicology Centre University of Saskatchewan Saskatoon SK CanadaAbstract Machine learning (ML) is increasingly considered the solution to environmental problems where limited or no physico‐chemical process understanding exists. But in supporting high‐stakes decisions, where the ability to explain possible solutions is key to their acceptability and legitimacy, ML can fall short. Here, we develop a method, rooted in formal sensitivity analysis, to uncover the primary drivers behind ML predictions. Unlike many methods for explainable artificial intelligence (XAI), this method (a) accounts for complex multi‐variate distributional properties of data, common in environmental systems, (b) offers a global assessment of the input‐output response surface formed by ML, rather than focusing solely on local regions around existing data points, and (c) is scalable and data‐size independent, ensuring computational efficiency with large data sets. We apply this method to a suite of ML models predicting various water quality variables in a pilot‐scale experimental pit lake. A critical finding is that subtle alterations in the design of some ML models (such as variations in random seed, functional class, hyperparameters, or data splitting) can lead to different interpretations of how outputs depend on inputs. Further, models from different ML families (decision trees, connectionists, or kernels) may focus on different aspects of the information provided by data, despite displaying similar predictive power. Overall, our results underscore the need to assess the explanatory robustness of ML models and advocate for using model ensembles to gain deeper insights into system drivers and improve prediction reliability.https://doi.org/10.1029/2024WR037398machine learningsensitivity analysisfeature selectionmodel randomnesswater quality
spellingShingle Banamali Panigrahi
Saman Razavi
Lorne E. Doig
Blanchard Cordell
Hoshin V. Gupta
Karsten Liber
On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
Water Resources Research
machine learning
sensitivity analysis
feature selection
model randomness
water quality
title On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
title_full On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
title_fullStr On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
title_full_unstemmed On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
title_short On Robustness of the Explanatory Power of Machine Learning Models: Insights From a New Explainable AI Approach Using Sensitivity Analysis
title_sort on robustness of the explanatory power of machine learning models insights from a new explainable ai approach using sensitivity analysis
topic machine learning
sensitivity analysis
feature selection
model randomness
water quality
url https://doi.org/10.1029/2024WR037398
work_keys_str_mv AT banamalipanigrahi onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis
AT samanrazavi onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis
AT lorneedoig onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis
AT blanchardcordell onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis
AT hoshinvgupta onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis
AT karstenliber onrobustnessoftheexplanatorypowerofmachinelearningmodelsinsightsfromanewexplainableaiapproachusingsensitivityanalysis