Optimization using Genetic Algorithm in Food Composition

Abstract

This paper aims to process daily food consumption based on the Guideline to Healthy Food or the principle of Healthy 4 Perfect 5 (H4P5). The Guideline to Healthy Food does not include the portion and size of daily food consumption, therefore it is perfected into the Guidelines to Balanced Nutrition (GBN). GBN is performed by consuming three servings of protein vegetables, three servings of animal protein, eight servings of carbohydrates, five servings of vegetables, five servings of fruits, and eight servings of water. Each part of the portion has a proportional size and dose based on GBN. The portions are measured by food composition using the optimization process using a Genetic Algorithm (GA). Optimization by a using Genetic Algorithm yielded three outputs, namely, first, the initial objective function (0) compared to the generative objective function (1) produces a value of 6.13: 4.72 (there is an increase in chromosomes, because, the smaller the value, the better the chromosome is). Second, probability mutation is 50% of mutation rate, crossover probability is 5%, the maximum generation is 36, the maximum fitness point is 0.240806, and the size of the population is 6. Third, optimization using the Genetic Algorithm in food composition produces one food model, one consumption model, and one composition model. Thus, the result of this research (food model, consumption model, and composition model) can be a recommendation for food consumption patterns (normal condition) and a reference for national food consumption policies.