Prediction of arterial hypertension through a logistic regression system

• Autores: Tesillo Gomez, Cynthia Mayumi, Escobar Arcaya, Yuri Alexander, León Gutierrez, Edwin Daniel
• Formato: artículo
• Fecha de Publicación: 2021
• Institución: Universidad La Salle
• Repositorio: Revistas - Universidad La Salle
• Lenguaje: español
• OAI Identifier: oai:ojs.revistas.ulasalle.edu.pe:article/44
• Enlace del recurso: https://revistas.ulasalle.edu.pe/innosoft/article/view/44; https://doi.org/10.48168/innosoft.s6.a44; https://purl.org/42411/s6/a44; https://n2t.net/ark:/42411/s6/a44
• Nivel de acceso: acceso abierto
• Materia: Arterial hypertension; Artificial Intelligence; Blood Pressure; Logistic Regression; Hipertensión arterial; Inteligencia Artificial; Regresión logística; Presión arterial

In Peru and the entire world, hypertension is a disease that can progress without showing any symptoms or these being very mild. You can have high blood pressure and not feel any manifestations, arterial hypertension is a serious public health problem in developing countries like ours: According to the 2017 Demographic and Family Health Survey Survey, although the prevalence of hypertension in people aged 15 years and over would have decreased from 14.8% in 2014 to 13.6%, it implies that more than 3 million Peruvians live with high blood pressure. For this reason, our goal is the rapid diagnosis of this silent disease. In the present work, the logistic regression system was used, for which there is a dataset of 5615 analyzed records. This article presents the possibility of detecting a disease such as high blood pressure based on artificial intelligence, since this evil has been increasing in the last years. For this reason, the objective is to quickly predict a possible diagnosis of arterial hypertension, for this, a dataset of 5615 records was analyzed in the Jupyter Notebook web application, establishing 9 input variables and 1 output, in addition, the logistic regression system was used, missing data treatments and outlaiers, graphs of variables, obtaining as a result an acceptable average precision of 87%.