Regresar

Application of data mining techniques for the prediction of tobacco and alcohol use in university students

Abstract:

Tobacco addiction is a social health problem, reaching 30% of the adult population worldwide. For health organizations, it is essential to have tools that allow them to apply care and prevention policies. A pilot study is proposed to discover patterns of characteristics on tobacco and alcohol poli-consumers in university students. Thus, the information was collected from 478 Ecuadorian young university students, obtaining records of 73 socio-demographic, psychosocial, health and consumption variables. A method of data mining is applied for prediction: pre-processing, sample analysis, application of algorithms and analysis of results. Seven variables with greater involvement in the prediction are obtained, which are validated in two experiments, applying decision trees and neural networks.