Development of a Genetic Algorithms Optimization Algorithm for a Nutritional Guidance Application

Abstract

Personalized easy to follow nutritional guidance is getting more important, since lifestyle related health problems are increasing. To gain a healthy balanced diet usually requires knowledge of a licensed nutritionist. There is a Fuzzy Expert System (FES) which applies knowledge of nutritionists, health data of an individual, personalized nutritional recommendation, and a meal diary with food composition data to balance a diet. FES generates a set of foods and beverages which should be altered in the diet with information on the direction and importance of the change. This paper presents a selection and a development of an optimization algorithm to be integrated with FES to provide easy to follow nutritional guidance. The selection process is carried out as a literature review. The development of selected Genetic Algorithms (GA) is carried out as an integrated part of Nutritional Guidance application, Nutri-Flow ® , since FES generates the search space, and is an important part of a Fitness Function of the optimization algorithm. The selection of the design parameters, are described and the test results are presented. Validation of the overall model is carried out with an expert analysis and comparison of the nutrient intake from the initial diet and recommended diet.