Método para identificar y pronosticar riesgo suicida perfiles de en adolescentes mediante técnicas de análisis de conglomerado y red neuronal artificial
Cargando...
Fecha
2019
Autores
Reyes-Ruiz, Lizeth
De La Hoz Granadillo, Efraín Javier
Carmona Alvarado, Farid Alejandro
Título de la revista
ISSN de la revista
Título del volumen
Editor
Sociedad Venezolana de Farmacología Clínica y Terapéutica
Resumen
En este trabajo de investigación, se presenta un método para
identificar y pronosticar perfiles de riesgo suicida en adolescentes
a partir del caso de estudio de estudiantes de secundaria
de escuela pública del departamento del Atlántico.
Como fundamento teórico se utilizó la literatura asociada a
la evaluación del riesgo suicida, el análisis de conglomerados
y las redes neuronales artificiales. Para lo anterior, se
tomó información relacionada con los factores de riesgo suicida
desesperanza, ideación, aislamiento y apoyo familiar de
una muestra de 119 estudiantes adolescentes de los grados
noveno al undécimo del departamento del Atlántico. Como
resultado se desarrolló un método para identificar, valorar y
pronosticar perfiles de riesgo suicida en adolescentes. Se
concluye que el análisis de conglomerados mostró condiciones
favorables para clasificar 4 perfiles característicos de
riesgo suicida y las redes neuronales artificiales capacidad
de pronosticar con una probabilidad del 95,5% de clasificación
correcta.
This article presents the profiles of suicide risk in 119 adolescent high school students of public schools of the department of Atlántico. As a theoretical basis, the literature associated with the evaluation of suicide risk, the analysis of conglomerates and the artificial neural networks were used. For the above, information was taken related to suicidal risk factors, hopelessness, ideation, isolation and family support. As a result, a method was developed to identify, assess and predict suicide risk profiles in adolescents. It is concluded that the cluster analysis showed favorable conditions to classify 4 characteristic profiles of suicide risk and artificial neural networks with a capacity to predict with a 95.5% probability of correct classification.
This article presents the profiles of suicide risk in 119 adolescent high school students of public schools of the department of Atlántico. As a theoretical basis, the literature associated with the evaluation of suicide risk, the analysis of conglomerates and the artificial neural networks were used. For the above, information was taken related to suicidal risk factors, hopelessness, ideation, isolation and family support. As a result, a method was developed to identify, assess and predict suicide risk profiles in adolescents. It is concluded that the cluster analysis showed favorable conditions to classify 4 characteristic profiles of suicide risk and artificial neural networks with a capacity to predict with a 95.5% probability of correct classification.
Descripción
Palabras clave
Riesgo suicida, Análisis de conglomerado, Redes neuronales artificiales, Suicide risk, Conglomerate analysis, Artificial neural network