Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots

Modern agriculture faces escalating challenges that demand predictive models balancing accuracy and interpretability to support informed decision-making. While precision agriculture has advanced through high-resolution remote sensing technologies capturing spectral, thermal, and soil data, interpret...

Full description

Saved in:
Bibliographic Details
Main Authors: J. Triana-Martinez, A. Álvarez-Meza, G. Castellanos-Dominguez
Format: Article
Language:English
Published: Elsevier 2025-09-01
Series:Results in Engineering
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2590123025027392
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849239327478906880
author J. Triana-Martinez
A. Álvarez-Meza
G. Castellanos-Dominguez
author_facet J. Triana-Martinez
A. Álvarez-Meza
G. Castellanos-Dominguez
author_sort J. Triana-Martinez
collection DOAJ
description Modern agriculture faces escalating challenges that demand predictive models balancing accuracy and interpretability to support informed decision-making. While precision agriculture has advanced through high-resolution remote sensing technologies capturing spectral, thermal, and soil data, interpreting these complex, nonlinear datasets remains a significant challenge for agronomists. This study introduces the TabNet-informed UMAP-based Local Biplot (UL-Biplot) framework, which combines TabNet's attention-based feature attribution with a Local Biplot technique from the widely recognized Uniform Manifold Approximation and Projection (UMAP). The method provides an intuitive representation of non-stationary and non-linear data patterns, enhancing both global and cluster-level explainability. Also, we highlight feature loadings based on TabNet's attention scores and evaluate the framework on a synthetic benchmark and two real-world datasets: a forage grasses trial for breeder score prediction and the WMBD wheat dataset for LAI estimation. Quantitative results demonstrate that our framework achieved an R2=0.79±0.01 for LAI and delivered more consistent performance for breeder scores (R2=0.77±0.01) relative to standard machine learning models. Moreover, TabNet's latent embeddings improved significantly in trustworthiness (from 0.94 to 0.98) and in class-neighbourhood preservation (0.047 versus 0.018 with conventional methods). The UL-Biplot also revealed distinct agronomic clusters associated with irrigation-induced greenness indices and moisture stress responses, providing actionable insights for site-specific crop management.
format Article
id doaj-art-bab3bd84d1dc4d1197fe546843c55d8d
institution Kabale University
issn 2590-1230
language English
publishDate 2025-09-01
publisher Elsevier
record_format Article
series Results in Engineering
spelling doaj-art-bab3bd84d1dc4d1197fe546843c55d8d2025-08-20T04:01:02ZengElsevierResults in Engineering2590-12302025-09-012710667210.1016/j.rineng.2025.106672Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local BiplotsJ. Triana-Martinez0A. Álvarez-Meza1G. Castellanos-Dominguez2Corresponding author.; Signal Processing and Recognition Group, Universidad Nacional de Colombia, Sede Manizales, 170001, ColombiaSignal Processing and Recognition Group, Universidad Nacional de Colombia, Sede Manizales, 170001, ColombiaSignal Processing and Recognition Group, Universidad Nacional de Colombia, Sede Manizales, 170001, ColombiaModern agriculture faces escalating challenges that demand predictive models balancing accuracy and interpretability to support informed decision-making. While precision agriculture has advanced through high-resolution remote sensing technologies capturing spectral, thermal, and soil data, interpreting these complex, nonlinear datasets remains a significant challenge for agronomists. This study introduces the TabNet-informed UMAP-based Local Biplot (UL-Biplot) framework, which combines TabNet's attention-based feature attribution with a Local Biplot technique from the widely recognized Uniform Manifold Approximation and Projection (UMAP). The method provides an intuitive representation of non-stationary and non-linear data patterns, enhancing both global and cluster-level explainability. Also, we highlight feature loadings based on TabNet's attention scores and evaluate the framework on a synthetic benchmark and two real-world datasets: a forage grasses trial for breeder score prediction and the WMBD wheat dataset for LAI estimation. Quantitative results demonstrate that our framework achieved an R2=0.79±0.01 for LAI and delivered more consistent performance for breeder scores (R2=0.77±0.01) relative to standard machine learning models. Moreover, TabNet's latent embeddings improved significantly in trustworthiness (from 0.94 to 0.98) and in class-neighbourhood preservation (0.047 versus 0.018 with conventional methods). The UL-Biplot also revealed distinct agronomic clusters associated with irrigation-induced greenness indices and moisture stress responses, providing actionable insights for site-specific crop management.http://www.sciencedirect.com/science/article/pii/S2590123025027392TabNetUMAPLocal BiplotFeature selectionInterpretabilityRemote sensing
spellingShingle J. Triana-Martinez
A. Álvarez-Meza
G. Castellanos-Dominguez
Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
Results in Engineering
TabNet
UMAP
Local Biplot
Feature selection
Interpretability
Remote sensing
title Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
title_full Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
title_fullStr Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
title_full_unstemmed Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
title_short Enhancing agricultural data interpretability and visualization with TabNet-driven feature extraction and Local Biplots
title_sort enhancing agricultural data interpretability and visualization with tabnet driven feature extraction and local biplots
topic TabNet
UMAP
Local Biplot
Feature selection
Interpretability
Remote sensing
url http://www.sciencedirect.com/science/article/pii/S2590123025027392
work_keys_str_mv AT jtrianamartinez enhancingagriculturaldatainterpretabilityandvisualizationwithtabnetdrivenfeatureextractionandlocalbiplots
AT aalvarezmeza enhancingagriculturaldatainterpretabilityandvisualizationwithtabnetdrivenfeatureextractionandlocalbiplots
AT gcastellanosdominguez enhancingagriculturaldatainterpretabilityandvisualizationwithtabnetdrivenfeatureextractionandlocalbiplots