Machine learning algorithms for soybean yield forecasting in the Brazilian Cerrado

Abstract

BACKGROUND: We evaluated different machine learning (ML) models for predicting soybean productivity up to 1 month in advance for the Matopiba agricultural frontier (States of Maranhão, Tocantins, Piauí, and Bahia). We collected meteorological data on the NASA-POWER platform and soybean yield on the SIDRA/IBGE base between 2008 and 2017. The ML models evaluated were random forest (RF), artificial neural networks, radial base support vector machines (SVM_RBF), linear model and polynomial regression. To assess the performance of the models, cross-validation was used, obtaining the value of precision by R 2 , accuracy by root mean square error (RMSE), and trend by the mean error of the estimate (EME). RESULTS:The results showed that the RF algorithm achieves the highest precision and accuracy, with R 2 of 0.81, RMSE of 176.93 kg ha −1 and trend (EME) of 1.99 kg ha −1 . On the other hand, the SVM_RBF algorithm showed the lowest performance, with R 2 of 0.74, RMSE of 213.58 kg ha −1 and EME of −15.06 kg ha −1 . The average yield values predicted by the models were within the expected range for the region, which has a historical average value of 2.730 kg ha −1 . CONCLUSION: All models had acceptable precision, accuracy and trend indices, which makes it possible to use all algorithms to be applied in the prediction of soybean crop yield, observing the particularities of the region to be studied, in addition to being a useful tool for agricultural planning and decision making in soy-producing regions such as the Brazilian Cerrado.

Comparison of Methods To Aggregate Climate Data To Predict Crop Yield: An Application to Soybean

“…This study showed that machine learning models using monthly climate averages were more accurate than linear regression to predict soybean yield, which consolidates evidence from numerous studies [9,22,23,27] in this area of research. In addition, RF models were less sensitive to the method chosen to aggregate climate data compared to LR models, whose performances strongly varied among the tested approaches.…”

“…Among ML techniques, RF was found to be one of the best algorithm in soybean yield prediction [9,22,23] as well as in other major crops including wheat [51] or maize [11]. Other articles report that other algorithms which have not been considered in our study such as neural networks [27] also perform well to forecast yield of soybean.…”

“…Standard regression models do not perform well when their inputs are highly correlated and do not always capture all the possible interactions between predictors [15]. Because of their greater flexibility, machine learning algorithms such as random forest (RF) [17], (extreme) gradient boosting [18,19], and deep learning [20] are now commonly used to predict yields of crops from climatic predictors [10,21] and often show better performances compared to traditional statistical methods such as linear regression, in particular to predict soybean yields [9,22,23]. 4 The use of statistical and machine learning methods raises a specific problem concerning the time step of the climate variables used as yield predictors.…”

Comparison of methods to aggregate climate data to predict crop yield: an application to soybean

“…Thanks to their greater flexibility, machine learning algorithms such as RF [29], (extreme) gradient boosting [37,38], and deep learning [39] are widely employed to predict yields of crops from climatic predictors [36,40] and often outperform traditional methods such as LR or process-based models, particularly in soybean yields predictions [8,41,42]. Among these techniques, RF was found to be one of the best algorithm in soybean yield prediction [8,41,42] as well as in other major crops including wheat [43] or maize [44]. Previous articles report that other algorithms which have not been considered in our study such as neural networks [12] also perform well to forecast yield of soybean.…”

Do Gridded Weather Datasets Provide High-Quality Data for Agroclimatic Research in Citrus Production in Brazil?

“…Additionally, this resource can be used to correct outliers and fill data gaps. Several works in the literature have explored using gridded datasets for crop yield prediction [28,[40][41][42]. However, only a few studies, such as Bai et al [38]; Bender and Sentelhas [43]; Battisti et al [44] and Duarte and Sentelhas [11], aimed to compare in situ and gridded data, which is essential for helping the modeler or decision-maker to choose which data sources to use on his/her crop yield prediction model.…”