Abstract
We investigate the predictive performance of two novel CNN-DNN machine learning ensemble models in predicting county-level corn yields across the US Corn Belt (12 states). The developed data set is a combination of management, environment, and historical corn yields from 1980 to 2019. Two scenarios for ensemble creation are considered: homogenous and heterogenous ensembles. In homogenous ensembles, the base CNN-DNN models are all the same, but they are generated with a bagging procedure to ensure they exhibit a certain level of diversity. Heterogenous ensembles are created from different base CNN-DNN models which share the same architecture but have different hyperparameters. Three types of ensemble creation methods were used to create several ensembles for either of the scenarios: Basic Ensemble Method (BEM), Generalized Ensemble Method (GEM), and stacked generalized ensembles. Results indicated that both designed ensemble types (heterogenous and homogenous) outperform the ensembles created from five individual ML models (linear regression, LASSO, random forest, XGBoost, and LightGBM). Furthermore, by introducing improvements over the heterogenous ensembles, the homogenous ensembles provide the most accurate yield predictions across US Corn Belt states. This model could make 2019 yield predictions with a root mean square error of 866 kg/ha, equivalent to 8.5% relative root mean square and could successfully explain about 77% of the spatio-temporal variation in the corn grain yields. The significant predictive power of this model can be leveraged for designing a reliable tool for corn yield prediction which will in turn assist agronomic decision makers.
Highlights
Accurate crop yield prediction is essential for agriculture production, as it can provide insightful information to farmers, agronomists, and other decision makers
Motivated by the high predictive performance of Convolutional neural networks (CNNs) and ensemble models in ecology (Cai et al, 2017; You et al, 2017; Yang et al, 2019; Khaki et al, 2020b; Shahhosseini et al, 2020, 2021), we propose a set of ensemble models created from multiple hybrid CNN-deep neural network (DNN) base learners for predicting county-level corn yields across US Corn Belt states
The CNN-recursive neural network (RNN) model developed by Khaki et al (2020b) presented test prediction errors of 988 kg/ha (15.74 bu/acre) and 1,107 kg/ha (17.64 bu/acre) for the test years 2017 and 2018, respectively, while the homogeneous ensemble model designed here resulted in test prediction errors of 874 kg/ha (13.93 bu/acre) and 992 kg/ha (15.8 bu/acre) for the test years 2017 and 2018, respectively
Summary
Accurate crop yield prediction is essential for agriculture production, as it can provide insightful information to farmers, agronomists, and other decision makers. This is not an easy task, as there is a myriad of variables that affect the crop yields, from genotypes, environment, and management decisions to technological advancements. The tools that are used to predict crop yields are mainly divided into simulation crop modeling and machine learning (ML). These models are usually utilized separately, there have been some recent studies to combine them toward improving prediction.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
