Selecting appropriate machine learning methods for digital soil mapping

Khaledian, Yones; Miller, Bradley A.

doi:10.1016/j.apm.2019.12.016

Cited by 300 publications

(138 citation statements)

References 80 publications

(96 reference statements)

Supporting

Mentioning

134

Contrasting

Unclassified

Order By: Relevance

“…132 The quality of our maps depends on the sampling design used to gather information, methods used, the models used to project point data across a 2-or 3-dimensional mapping product, and the skill and training of those who create the maps. 98,[133][134][135] Therefore, additional research is needed to investigate best soil mapping practices including data collection, analysis, manipulation, and display. 136…”

Section: Future Needs In Soil Pollution and Human Healthmentioning

confidence: 99%

Soil and Human Health: Current Status and Future Needs

Brevik

Slaughter

Singh

et al. 2020

Air, Soil and Water Research

199

109

View full text Add to dashboard Cite

Soil influences human health in a variety of ways, with human health being linked to the health of the soil. Historically, emphasis has been placed on the negative impacts that soils have on human health, including exposures to toxins and pathogenic organisms or the problems created by growing crops in nutrient-deficient soils. However, there are a number of positive ways that soils enhance human health, from food production and nutrient supply to the supply of medications and enhancement of the immune system. It is increasingly recognized that the soil is an ecosystem with a myriad of interconnected parts, each influencing the other, and when all necessary parts are present and functioning (ie, the soil is healthy), human health also benefits. Despite the advances that have been made, there are still many areas that need additional investigation. We do not have a good understanding of how chemical mixtures in the environment influence human health, and chemical mixtures in soil are the rule, not the exception. We also have sparse information on how most chemicals react within the chemically and biologically active soil ecosystem, and what those reactions mean for human health. There is a need to better integrate soil ecology and agronomic crop production with human health, food/nutrition science, and genetics to enhance bacterial and fungal sequencing capabilities, metagenomics, and the subsequent analysis and interpretation. While considerable work has focused on soil microbiology, the macroorganisms have received much less attention regarding links to human health and need considerable attention. Finally, there is a pressing need to effectively communicate soil and human health connections to our broader society, as people cannot act on information they do not have. Multidisciplinary teams of researchers, including scientists, social scientists, and others, will be essential to move all these issues forward.

show abstract

Section: Future Needs In Soil Pollution and Human Healthmentioning

confidence: 99%

Soil and Human Health: Current Status and Future Needs

Brevik

Slaughter

Singh

et al. 2020

Air, Soil and Water Research

199

109

View full text Add to dashboard Cite

show abstract

“…In this study, four common performance metrics [98], namely root mean squared error (RMSE), normalized root mean squared error (nRMSE), coefficient of determination (R 2 ), and Ratio of Performance to InterQuartile distance (RPIQ) were used. RMSE indicates the accuracy of the model prediction.…”

Section: Statistical Evaluationmentioning

confidence: 99%

Improving the Spatial Prediction of Soil Organic Carbon Content in Two Contrasting Climatic Regions by Stacking Machine Learning Models and Rescanning Covariate Space

et al. 2020

View full text Add to dashboard Cite

Understanding the spatial distribution of soil organic carbon (SOC) content over different climatic regions will enhance our knowledge of carbon gains and losses due to climatic change. However, little is known about the SOC content in the contrasting arid and sub-humid regions of Iran, whose complex SOC–landscape relationships pose a challenge to spatial analysis. Machine learning (ML) models with a digital soil mapping framework can solve such complex relationships. Current research focusses on ensemble ML models to increase the accuracy of prediction. The usual ensemble method is boosting or weighted averaging. This study proposes a novel ensemble technique: the stacking of multiple ML models through a meta-learning model. In addition, we tested the ensemble through rescanning the covariate space to maximize the prediction accuracy. We first applied six state-of-the-art ML models (i.e., Cubist, random forests (RF), extreme gradient boosting (XGBoost), classical artificial neural network models (ANN), neural network ensemble based on model averaging (AvNNet), and deep learning neural networks (DNN)) to predict and map the spatial distribution of SOC content at six soil depth intervals for both regions. In addition, the stacking of multiple ML models through a meta-learning model with/without rescanning the covariate space were tested and applied to maximize the prediction accuracy. Out of six ML models, the DNN resulted in the best modeling accuracies, followed by RF, XGBoost, AvNNet, ANN, and Cubist. Importantly, the stacking of models indicated a significant improvement in the prediction of SOC content, especially when combined with rescanning the covariate space. For instance, the RMSE values for SOC content prediction of the upper 0–5 cm of the soil profiles of the arid site and the sub-humid site by the proposed stacking approaches were 17% and 9% respectively, less than that obtained by the DNN models—the best individual model. This indicates that rescanning the original covariate space by a meta-learning model can extract more information and improve the SOC content prediction accuracy. Overall, our results suggest that the stacking of diverse sets of models could be used to more accurately estimate the spatial distribution of SOC content in different climatic regions.

show abstract

“…On the other hand, there is evidence that ANN and DT models are better than LR for predicting soil classes. An evaluation of the prediction models of soil properties (Khaledian and Miller, 2020) concluded that there is no one single correct learning algorithm. However, certain algorithms are more appropriate than others considering the purpose of the mapping.…”

Section: Classifiersmentioning

confidence: 99%

Digital soil class mapping in Brazil: a systematic review

Coelho

Giasson

Campos

et al. 2021

Sci. agric. (Piracicaba, Braz.)

View full text Add to dashboard Cite

In Brazil several digital soil class mapping studies were carried out from 2006 onwards to maximize the use of existing maps and information and to provide estimates for wider areas. However, there is no consensus on which methods have produced superior results in the predictive value of soil maps. This study conducts a systematic review of digital soil class mapping in Brazil and aims to analyze the factors which can improve the accuracy of digital soil class maps. Data from 334 digital soil class mapping studies were grouped and analyzed by Student's t-test, Wilcoxon-Mann-Whitney test and Kruskal-Wallis test. When conventional maps were used for validation, the studies showed average values of 63 % and when field samples were used, 56 % for Overall Accuracy. Studies compatible with the Planimetric Cartographic Accuracy Standard for Digital Cartographic Products (PEC-PCD) averaged between 4 % and 15 % higher accuracy than those of the incompatible group. There seems to be no evidence that increasing the number of variables and samples results in more accurate soil map prediction, but studies using variables related to four soil-forming factors enhanced accuracy. From a density of 0.08 MU km-2 and upwards, it became more difficult for studies to obtain greater accuracy. Artificial neural network classifiers and Decision Tree models seem to be producing more accurate digital soil class maps.

show abstract

Selecting appropriate machine learning methods for digital soil mapping

Cited by 300 publications

References 80 publications

Soil and Human Health: Current Status and Future Needs

Soil and Human Health: Current Status and Future Needs

Improving the Spatial Prediction of Soil Organic Carbon Content in Two Contrasting Climatic Regions by Stacking Machine Learning Models and Rescanning Covariate Space

Digital soil class mapping in Brazil: a systematic review

Contact Info

Product

Resources

About