Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data

Abstract This paper addresses some of the challenges in automating mineral mapping from CRISM hyperspectral data using two ML algorithms, namely, Random Forest (RF) and Gradient Boosted Trees (GBTree) algorithms. An interpretable framework using tree‐ensemble classification and SHapely Additive exPl...

Full description

Saved in:
Bibliographic Details
Main Authors: Sandeepan Dhoundiyal, Moni Shankar Dey, Shashikant Singh, Pattathal V. Arun, Guneshwar Thangjam, Alok Porwal
Format: Article
Language:English
Published: Wiley 2025-06-01
Series:Journal of Geophysical Research: Machine Learning and Computation
Subjects:
Online Access:https://doi.org/10.1029/2024JH000391
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850167637497085952
author Sandeepan Dhoundiyal
Moni Shankar Dey
Shashikant Singh
Pattathal V. Arun
Guneshwar Thangjam
Alok Porwal
author_facet Sandeepan Dhoundiyal
Moni Shankar Dey
Shashikant Singh
Pattathal V. Arun
Guneshwar Thangjam
Alok Porwal
author_sort Sandeepan Dhoundiyal
collection DOAJ
description Abstract This paper addresses some of the challenges in automating mineral mapping from CRISM hyperspectral data using two ML algorithms, namely, Random Forest (RF) and Gradient Boosted Trees (GBTree) algorithms. An interpretable framework using tree‐ensemble classification and SHapely Additive exPlanations (SHAP) is implemented to interpret the model decisions. SHAP explanations quantify the influence of diagnostic absorption features, demonstrating that classifiers rely on physically significant spectral features, rather than artifacts. Novel metrics: “Physically Significant Precision,” “Physically Significant Recall,” and “Physically Significant F‐measure” quantify the classifier's expected performance on unseen data. RF outperforms GBTree, and thus is used to develop a novel framework for mineral mapping from CRISM data, demonstrated on five CRISM datacubes. This amalgamation of Random Forest and SHAP addresses limitations associated with existing CRISM classification methods, offering stability during training, reduced manual intervention, and interpretability while achieving a Kappa (κ) of 0.91 over the CRISM Machine Learning Toolkit's mineral data set with ∼470,000 labeled spectra.
format Article
id doaj-art-8cbec6a79d284efaad87ba1aea683ea2
institution OA Journals
issn 2993-5210
language English
publishDate 2025-06-01
publisher Wiley
record_format Article
series Journal of Geophysical Research: Machine Learning and Computation
spelling doaj-art-8cbec6a79d284efaad87ba1aea683ea22025-08-20T02:21:10ZengWileyJournal of Geophysical Research: Machine Learning and Computation2993-52102025-06-0122n/an/a10.1029/2024JH000391Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral DataSandeepan Dhoundiyal0Moni Shankar Dey1Shashikant Singh2Pattathal V. Arun3Guneshwar Thangjam4Alok Porwal5Centre of Studies in Resources Engineering Indian Institute of Technology Bombay Mumbai IndiaCentre of Studies in Resources Engineering Indian Institute of Technology Bombay Mumbai IndiaCentre of Studies in Resources Engineering Indian Institute of Technology Bombay Mumbai IndiaComputer Science and Engineering Group Indian Institute of Information Technology Sri City Sri City IndiaSchool Earth & Planetary Sciences National Institute of Science Education and Research Bhubhaneshwar IndiaCentre of Studies in Resources Engineering Indian Institute of Technology Bombay Mumbai IndiaAbstract This paper addresses some of the challenges in automating mineral mapping from CRISM hyperspectral data using two ML algorithms, namely, Random Forest (RF) and Gradient Boosted Trees (GBTree) algorithms. An interpretable framework using tree‐ensemble classification and SHapely Additive exPlanations (SHAP) is implemented to interpret the model decisions. SHAP explanations quantify the influence of diagnostic absorption features, demonstrating that classifiers rely on physically significant spectral features, rather than artifacts. Novel metrics: “Physically Significant Precision,” “Physically Significant Recall,” and “Physically Significant F‐measure” quantify the classifier's expected performance on unseen data. RF outperforms GBTree, and thus is used to develop a novel framework for mineral mapping from CRISM data, demonstrated on five CRISM datacubes. This amalgamation of Random Forest and SHAP addresses limitations associated with existing CRISM classification methods, offering stability during training, reduced manual intervention, and interpretability while achieving a Kappa (κ) of 0.91 over the CRISM Machine Learning Toolkit's mineral data set with ∼470,000 labeled spectra.https://doi.org/10.1029/2024JH000391hyperspectral remote sensingCRISMmineral mappingMarsExplainbale AI
spellingShingle Sandeepan Dhoundiyal
Moni Shankar Dey
Shashikant Singh
Pattathal V. Arun
Guneshwar Thangjam
Alok Porwal
Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
Journal of Geophysical Research: Machine Learning and Computation
hyperspectral remote sensing
CRISM
mineral mapping
Mars
Explainbale AI
title Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
title_full Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
title_fullStr Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
title_full_unstemmed Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
title_short Explainable Machine Learning for Mapping Minerals From CRISM Hyperspectral Data
title_sort explainable machine learning for mapping minerals from crism hyperspectral data
topic hyperspectral remote sensing
CRISM
mineral mapping
Mars
Explainbale AI
url https://doi.org/10.1029/2024JH000391
work_keys_str_mv AT sandeepandhoundiyal explainablemachinelearningformappingmineralsfromcrismhyperspectraldata
AT monishankardey explainablemachinelearningformappingmineralsfromcrismhyperspectraldata
AT shashikantsingh explainablemachinelearningformappingmineralsfromcrismhyperspectraldata
AT pattathalvarun explainablemachinelearningformappingmineralsfromcrismhyperspectraldata
AT guneshwarthangjam explainablemachinelearningformappingmineralsfromcrismhyperspectraldata
AT alokporwal explainablemachinelearningformappingmineralsfromcrismhyperspectraldata