Predictive models and food security

The study conducts a comprehensive review of the use of predictive models to strengthen food security in Latin America, with an emphasis on Colombia. It analyzes 65 studies published between 2014 and 2024 that apply algorithms such as Random Forests and Neural Networks to anticipate agricultural yields, manage climate risks, and optimize resources. The results show that agricultural, climatic, and technological variables predominate, while socioeconomic and nutritional dimensions are underrepresented. The article highlights the need to integrate multidimensional data to develop more robust models that strengthen the resilience and sustainability of food systems in the region.