MLFoodProcessing
MLFoodProcessing is a machine learning toolkit accompanying the research paper Machine Learning Prediction of the Degree of Food Processing by Menichetti, Ravandi, Mozaffarian, and Barabasi. It implements FoodProX and FPro, computational models that predict the degree of food processing for food items based on their nutritional and ingredient composition. The repository provides scripts for training, evaluating, and applying these models, enabling researchers to classify foods and analyze dietary datasets. Key use cases include nutritional epidemiology, food system research, and public health studies investigating the relationship between ultra-processed food consumption and health outcomes. The software combines Python, MATLAB, and R code, requiring Python 3.6.10 with packages such as scikit-learn, pandas, shap, networkx, and imbalanced-learn. A standard desktop with 16 GB RAM suffices for minimal use, while 64 GB RAM and 16 cores are recommended for optimal performance. Installation typically takes less tha