Segmenting the eating behaviour of university students using the K-means algorithm

Abstract

Universities that teach arts education do not only teach how to play an instrument or conduct musical ensembles; they are agents of change in eating behaviour for the praxis of teaching and dissemination of healthy education. The objective of the research was to segment the eating behaviour of students of the artistic education-music speciality of the National University of Education "Enrique Guzmán y Valle" by applying the K-means algorithm. To do this, the methodology consisted of understanding the problem, understanding the data collected, preparing the data, modelling and evaluating the model. For modelling, the free software Weka was used through the K-means clustering technique on a data matrix of 148 instances with forty-three nominal variables collected online based on an instrument designed and validated to assess eating behaviour in university students. Two was determined to be the optimal clustering for eating behaviour in university students, using the elbow method, with a distribution of 49% for the first cluster and 51% for the second cluster. The results of the study population showed that the eating behaviour of university students is adequate.