Food access of poor households in Mexico: a classification tree application

Poverty and lack of food access are critical problems in developing countries, affecting more than 20 % of the population in Mexico. This research aimed to analyze the factors influencing food access in Mexican households and to develop a predictive model. The machine learning technique, known as cl...

Full description

Bibliographic Details
Authors: Vázquez-Ayanegui, Fernando Elester, Hernández-Ortiz, Juan, Valdivia-Alcalá, Ramón, Botello-Aguillón, César, Toledo-Ruiz, Federico Augusto
Format: article
Status:Published version
Publication Date:2024
Country:México
Institution:UNIVERSIDAD AUTÓNOMA CHAPINGO
Repository:Revista de Geografía Agrícola
Language:Spanish
OAI Identifier:oai:ojs2.revistas.chapingo.mx:article/606
Online Access:https://revistas.chapingo.mx/geografia/article/view/606
Access Level:Open access
Keyword:Políticas sociales
carencias monetarias
líneas de bienestar
canasta alimentaria
hambre
Social policies
monetary deprivation
monetary deprivation,
hunger
Description
Summary:Poverty and lack of food access are critical problems in developing countries, affecting more than 20 % of the population in Mexico. This research aimed to analyze the factors influencing food access in Mexican households and to develop a predictive model. The machine learning technique, known as classification tree, was used and the results were compared with those of a logit model, frequently used in the literature. In terms of accuracy, the classification tree outperformed the logit model inidentifying households at risk of food insecurity (0.6039 vs. 0.5402) and provided a visual interpretation of the results. The findings suggest that households living in poverty, located in urban areas, with more than three members or without basic education, should be prioritized in social policies, since they are more likely to face food access problems in Mexico