Anna Kiss
Szent István University, Hungary
Title: How Much Is Enough? - Evidence Based Stochastic Optimisation Of Hungarian Nutrition Structure
Biography
Biography: Anna Kiss
Abstract
Background: the prevalence of obesity among the Hungarian adult population is one of the highest in Europe, being one of the main factors of mortality.
Motivation: A frequent counter-argument in framework of debates on modification of nutrition structure of population is the high cost of changing to more “healthier” nutritional patters. This is an extremely important problem in a middle-income country, where the food-related expenses are as high as 40% of disposable income of households. Our goal has been to determine the characteristic features of current nutrition structure, cost of it, and the cost of optimised nutrition structure.
Methodology: In framework of a preliminary study of a national-wide survey, face-to-face interviews have been carried out to determine the food consumption structure of 80 Hungarian households in two non-consecutive days, offering information on nutrition of nearly 200 respondents. The sample was distorted, because the dwellers of capital of Hungary, and the intelligentsia have been over-represented in it, but could furnish reliable information on consumption-structure of middle-, and middle-upper class of the society.
Dataset has been analysed by different sophisticated artificial intelligence methods (machine learning algorithms), with purpose of obtaining an optimal classification of most characteristic food consumption patterns.
On base of patterns, offering the best accuracy, as well as internationally accepted data-bases on recommended nutrition intake, taking into consideration the physical activity as well as demographic characteristics of the sample and the actual procuration prices of different products, applying the Linear Programming algorithm of Lindo® Systerms a recommended nutrition structure has been developed for each pattern.
Results: comparative analysis of actual and optimised nutrition intake values highlights the false argument of high cost of healthy nutrition. This fact opens new frontiers for the tailor-made mobile applications, offering a suitable help for consumers of healthy, easy, cost-effective and sustainable food choice.