Predictive Modeling of Glyphosate Contamination Utilizing Random Forest Classifiers
Critical Debates in Humanities, Science and Global Justice · 2024
DOI: 10.68709/001c.123526 ↗Kurz gefasst
Fachartikel in der Zeitschrift Critical Debates in Humanities, Science and Global Justice, 2024. Verfasst von Jonathan Yang et al. und 1 weiteren Autor:innen. Die Arbeit nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Glyphosate, also known as “Roundup,” is a common herbicide used in various agricultural settings. Although the extent of this glyphosate’s harm to human health is subject to debate, this study seeks to use machine learning and data analysis to predict glyphosate contamination levels across the United States and help guide environmental testing and research. We utilized publicly available pesticide data provided by the United States Geological Survey (USGS) and county-level environmental and socioeconomic indicators with the EJScreen API. Two Random Forest Classifier models were trained on EJScreen environmental and socioeconomic data to predict glyphosate levels in counties across the U.S. that did report usage amounts. Although Pearson correlation coefficients revealed minimal correlation between glyphosate usage and EJScreen features, the models were able to predict glyphosate usage with high accuracy and precision scores of around 75%. This suggests that the models can capture complex, non-linear interactions among multiple features. Out of counties that did not report glyphosate usage, Roanoke and Covington, VA, were predicted by both models to have high glyphosate levels. More broadly, this study highlights the potential of machine learning in pesticide detection and management, offering valuable applications for national and state-level environmental testing. Integrating machine learning techniques can revolutionize environmental health decision-making, providing scalable and replicable models for other contaminants. These models can uncover trends and interactions in data that conventional methods may miss, leading to a more comprehensive understanding of environmental contamination and guiding targeted testing and mitigation strategies.
Zitieren
APA
al., J. Y. E., & Yang, J. (2024). Predictive Modeling of Glyphosate Contamination Utilizing Random Forest Classifiers. Critical Debates in Humanities, Science and Global Justice. https://doi.org/10.68709/001c.123526
Vancouver
al. JYE, Yang J. Predictive Modeling of Glyphosate Contamination Utilizing Random Forest Classifiers. Critical Debates in Humanities, Science and Global Justice. 2024. doi:10.68709/001c.123526
BibTeX
@article{al2024,
author = {al., Jonathan Yang et and Yang, Jonathan},
title = {Predictive Modeling of Glyphosate Contamination Utilizing Random Forest Classifiers},
journal = {Critical Debates in Humanities, Science and Global Justice},
year = {2024},
doi = {10.68709/001c.123526}
}
EUVIMED