A MACHINE LEARNING APPROACH TO ELUCIDATE INTERACTIONS OF FOOD CONSTITUENTS WITH NOVEL CAN COATINGS TO PREDICT PRODUCT SAFETY AND SHELF LIFE

The overarchinggoal is to reduce corrosion failure mechanisms inlight metal packaging lined with non-BPA coatings, thereby improving food safety, reducing food waste, and extending product shelf life. To achieve this goal, we set three objectives: (1) determine the corrosivity of food ingredients that are in contact with systematically designed food can coatings, (2) determine the effect of and mechanism by which coating properties influence the migration of food components through the food can coatings, and(3) select and implement a suitable machine learning approach for food-packaging interactions and their corresponding coating performance.