Machine learning-based approaches to identify key drivers and predict severe food insecurity in Africa

Extracting knowledge and discovering insights from hidden patterns in food insecurity data through machine learning algorithms remains limited. Therefore, this study employed a variety of machine learning algorithms, including linear regression, ridge regression, LASSO regression, elastic net regression, support vector regression, decision tree regression, random forest regression, gradient boosting regression, extrem gradient bosting (XGBoost) regression, k-nearest neighbors (KNN) regression, and neural networks (NN) to investigate the predictive associations between key drivers and severe food insecurity across African countries from 2015 to 2021. Model performance was improved through hyperparameter tuning via GridSearchCV, with XGBoosting outperforming all other models. Feature importance was assessed using permutation importance techniques, while recursive feature elimination (RFE) with cross-validation identified the most influential national-level predictors. SHapley Additive exPlanations (SHAP) further revealed the direction and magnitude of each feature’s effect. Results show that higher national-level of inflation, climate change, unemployment, malaria incidence, GHG emissions, caloric loss, and cereal import dependency are associated with higher predicted rates of severe food insecurity, whereas a higher livestock production, cereal crop production, investment inflows, dietary energy supply, dietary protein supply, electricity access, GDP per capita, and political stability are associated with lower predicted rates. Among these drivers, unemployment, livestock production, greenhouse gas (GHG) emissions, and caloric losses are the most influential drivers in predicting severe food insecurity across Africa, listed here in priority order. In light of these findings, the study recommends that policymakers prioritize interventions targeting these top-ranked factors when formulating and implementing national strategies to effectively mitigate severe food insecurity across African countries.