Recommending healthy meal plans using a many-objective optimization approach

Abstract

In today’s world, healthy eating is a very important issue affecting a large proportion of the population. In this study, we propose to extend the classical diet problem formulated by Stigler in 1945 and model it as a many-objective optimization problem. In our model, the objectives are inspired from the foraging behaviour of animals and take into account user preferences as well as cost, while satisfying the recommended daily nutrient intake constraints for a user’s gender and age group as defined by the USDA. Inspired by the animal foraging theory, the proposed many-objective diet problem has several objectives, such as: minimize cost, maximize user’s liked foods, maximize availability of ingredients, minimize meal preparation times, maximize variety in meal plans. For making meal plan recommendations, a database containing complete recipes (like "cream of mushroom soup"), as opposed to individual food items (like mushroom), is used. Therefore, the recommendations become more realistic meal plans, which in turn make them more applicable by the users. As far as the authors know, this is the first study that models the diet problem as a many-objective and multi-constraint optimization problem and solves it using modern meta-heuristics such as many-objective evolutionary or ant-colony algorithms to make healthy meal recommendations. Thus, the study can have an academic impact as well as a social one by helping people stay healthy through eating healthy meals.