Personalidad y Malestar PsicológicoAplicación de un Modelo de Redes Neuronales
ISSN: 1135-3848
Year of publication: 2016
Volume: 1
Issue: 41
Pages: 28-38
Type: Article
More publications in: Revista iberoamericana de diagnóstico y evaluación psicológica
Abstract
Feature selection using artificial neural networks is a non-linear technique useful for the study of a field of knowledge by means of its simplification. The goal of this work has been to prove the utility of this methodology on reducing the complexity of the study of the psychological discomfort. A sample of 81 students completed a battery of personality tests including personality traits, affect, motivation, expectancies and personality disorders. We created four different networks that used the personality traits as predictors of the symptomatology associated to the perceived discomfort: somatizations, anxiety, depression and social dysfunction. Results show the neural network utility to detect a specific collection of different nature predictors for each symptomatic group.