Explainable machine learning driven nutrient recommendation for maize production in Malawi

Addressing the persistent challenge of low maize productivity in Malawi requires spatially explicit and nutrient-specific fertilizer recommendations that align with soil heterogeneity and economic constraints. Existing blanket recommendations often overlook localized variability in soil properties,...

Full description

Bibliographic Details
Main Authors: Liben, Feyera, Kihara, Job, Abera, Wuletawu, AbebeMesfin, Tewodrofins, Homann-kee Tui, Sabine, Munthali, Moses, Munthali, Chandiona, Nalivata, Patson, Mzumara, Edward, Tuimene, Lulseged
Format: Abstract
Language:Inglés
Published: 2025
Subjects:
Online Access:https://hdl.handle.net/10568/179331
_version_ 1855526821333827584
author Liben, Feyera
Kihara, Job
Abera, Wuletawu
AbebeMesfin, Tewodrofins
Homann-kee Tui, Sabine
Munthali, Moses
Munthali, Chandiona
Nalivata, Patson
Mzumara, Edward
Tuimene, Lulseged
author_browse AbebeMesfin, Tewodrofins
Abera, Wuletawu
Homann-kee Tui, Sabine
Kihara, Job
Liben, Feyera
Munthali, Chandiona
Munthali, Moses
Mzumara, Edward
Nalivata, Patson
Tuimene, Lulseged
author_facet Liben, Feyera
Kihara, Job
Abera, Wuletawu
AbebeMesfin, Tewodrofins
Homann-kee Tui, Sabine
Munthali, Moses
Munthali, Chandiona
Nalivata, Patson
Mzumara, Edward
Tuimene, Lulseged
author_sort Liben, Feyera
collection Repository of Agricultural Research Outputs (CGSpace)
description Addressing the persistent challenge of low maize productivity in Malawi requires spatially explicit and nutrient-specific fertilizer recommendations that align with soil heterogeneity and economic constraints. Existing blanket recommendations often overlook localized variability in soil properties, climate, and nutrient responses, limiting their effectiveness. This study aimed to (1) evaluate and compare the predictive performance of multiple machine learning algorithms for maize yield estimation in Malawi; (2) identify the most important yield-determining features through recursive feature elimination (RFECV) and SHAP-based interpretation; (3) examine interaction effects between key nutrient inputs and environmental variables using two-dimensional partial dependence plots; and (4) translate model outputs into site-specific nutrient management insights for precision agronomy in smallholder maize systems. A Light Gradient Boosting Machine (LightGBM) model was selected based on performance and interpretability and trained on 20 selected features including nutrients management, soil, climate, topography, and hydrological covariates. Hyperparameter optimization and SHAP analysis enhanced both accuracy and model transparency. We generated fine-scale (250 m resolution) maize fertilizer recommendations for N, P, K, S, and Zn using the optimized LightGBM model. Results revealed substantial spatial variation in nutrient requirements, with nitrogen and phosphorus showing the widest ranges in both agronomic and economic rates. The economic optimum rates were slightly lower than agronomic ones, highlighting the importance of cost-benefit considerations in fertilizer planning. Spatial patterns also indicated that highly weathered soils in southern and central Malawi often required higher inputs, while nutrient-rich areas in the north exhibited lower optimal rates. These findings support a shift from national blanket recommendations to precision-guided input planning using high-resolution geospatial data and data-driven models. Keywords: Artificial intelligence; machine learning; nutrient; maize; LightGBM; SHAP
format Abstract
id CGSpace179331
institution CGIAR Consortium
language Inglés
publishDate 2025
publishDateRange 2025
publishDateSort 2025
record_format dspace
spelling CGSpace1793312025-12-31T02:08:31Z Explainable machine learning driven nutrient recommendation for maize production in Malawi Liben, Feyera Kihara, Job Abera, Wuletawu AbebeMesfin, Tewodrofins Homann-kee Tui, Sabine Munthali, Moses Munthali, Chandiona Nalivata, Patson Mzumara, Edward Tuimene, Lulseged maize machine learning artificial intelligence fertilizer application Addressing the persistent challenge of low maize productivity in Malawi requires spatially explicit and nutrient-specific fertilizer recommendations that align with soil heterogeneity and economic constraints. Existing blanket recommendations often overlook localized variability in soil properties, climate, and nutrient responses, limiting their effectiveness. This study aimed to (1) evaluate and compare the predictive performance of multiple machine learning algorithms for maize yield estimation in Malawi; (2) identify the most important yield-determining features through recursive feature elimination (RFECV) and SHAP-based interpretation; (3) examine interaction effects between key nutrient inputs and environmental variables using two-dimensional partial dependence plots; and (4) translate model outputs into site-specific nutrient management insights for precision agronomy in smallholder maize systems. A Light Gradient Boosting Machine (LightGBM) model was selected based on performance and interpretability and trained on 20 selected features including nutrients management, soil, climate, topography, and hydrological covariates. Hyperparameter optimization and SHAP analysis enhanced both accuracy and model transparency. We generated fine-scale (250 m resolution) maize fertilizer recommendations for N, P, K, S, and Zn using the optimized LightGBM model. Results revealed substantial spatial variation in nutrient requirements, with nitrogen and phosphorus showing the widest ranges in both agronomic and economic rates. The economic optimum rates were slightly lower than agronomic ones, highlighting the importance of cost-benefit considerations in fertilizer planning. Spatial patterns also indicated that highly weathered soils in southern and central Malawi often required higher inputs, while nutrient-rich areas in the north exhibited lower optimal rates. These findings support a shift from national blanket recommendations to precision-guided input planning using high-resolution geospatial data and data-driven models. Keywords: Artificial intelligence; machine learning; nutrient; maize; LightGBM; SHAP 2025-12-22 2025-12-30T15:06:53Z 2025-12-30T15:06:53Z Abstract https://hdl.handle.net/10568/179331 en Open Access application/vnd.openxmlformats-officedocument.wordprocessingml.document Liben, F.; Kihara, J.; Abera, W.; AbebeMesfin, T.; Homann-kee Tui, S.; Munthali, M..; Munthali, C.; Nalivata, P..; Mzumara, E.; Tuimene, L. (2025) Explainable machine learning driven nutrient recommendation for maize production in Malawi. 1 p.
spellingShingle maize
machine learning
artificial intelligence
fertilizer application
Liben, Feyera
Kihara, Job
Abera, Wuletawu
AbebeMesfin, Tewodrofins
Homann-kee Tui, Sabine
Munthali, Moses
Munthali, Chandiona
Nalivata, Patson
Mzumara, Edward
Tuimene, Lulseged
Explainable machine learning driven nutrient recommendation for maize production in Malawi
title Explainable machine learning driven nutrient recommendation for maize production in Malawi
title_full Explainable machine learning driven nutrient recommendation for maize production in Malawi
title_fullStr Explainable machine learning driven nutrient recommendation for maize production in Malawi
title_full_unstemmed Explainable machine learning driven nutrient recommendation for maize production in Malawi
title_short Explainable machine learning driven nutrient recommendation for maize production in Malawi
title_sort explainable machine learning driven nutrient recommendation for maize production in malawi
topic maize
machine learning
artificial intelligence
fertilizer application
url https://hdl.handle.net/10568/179331
work_keys_str_mv AT libenfeyera explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT kiharajob explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT aberawuletawu explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT abebemesfintewodrofins explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT homannkeetuisabine explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT munthalimoses explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT munthalichandiona explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT nalivatapatson explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT mzumaraedward explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi
AT tuimenelulseged explainablemachinelearningdrivennutrientrecommendationformaizeproductioninmalawi