Implementation of the ID3 algorithm for the generation of a decision tree with food health data from the State of Guerrero, Mexico

Abstract

Mexico faces a significant public health problem related to the eating habits of the population. More than 50% of the world's obese people live in 10 countries, including Mexico (Ng et al., 2014). In 2014, Mexico had the highest obesity prevalence rate in Latin America (PAHO, 2016a). The prevalence of obesity in adults in the state of Guerrero in 2012 was 71.5% (Barquera, 2012) The objective of this research was to generate a diagnosis of these conditions in both urban and rural locations in the seven regions of the state of Guerrero. The information was analyzed using an information system called SIOB, which is a technological tool developed to collect information from the population with a sample of 23,000 people, users of the public health sector of the state of Guerrero. We used the Weka tool of the University of Waikato (Eibe et al., 201d6) to generate a visible decision tree that shows the prevalence of these conditions in the state of Guerrero. It is considered that these visualization tools can help in the implementation of public policies that contribute to the prevention of these chronic degenerative diseases.