Identifying Risk Factors and Predicting Food Security Status Using Supervised Machine Learning Techniques

This study investigates the identification of risk factors and prediction of food security status utilizing various supervised machine learning algorithms, including Decision Trees, Random Forest, Multi-layer Perceptron (MLP), and XGBoost. The objective is to develop a robust predictive model that accurately determines the Food Security Index as Low, Moderate, or High based on diverse indicators. Using a comprehensive dataset encompassing socio-economic, environmental, and demographic variables, the research employs these machine learning techniques to analyze and classify the food security status of different regions or households. The Decision Tree algorithm provides an initial understanding of feature importance, while Random Forest harnesses the power of ensemble learning for enhanced accuracy. Additionally, the MLP neural network and XGBoost algorithm are employed to capture complex nonlinear relationships and boost predictive performance. Evaluation metrics such as accuracy, precision, recall, and F1-score are utilized to assess the models' performance. The results showcase the effectiveness of these supervised learning methods in accurately predicting food security status. Moreover, the study identifies key risk factors contributing to different levels of food security, providing valuable insights for policymakers and stakeholders to target interventions and improve food access and availability.. 

 Keywords: Decision Trees, Random Forest, Multi-layer Perceptron (MLP), and XGBoost.