<abstract xmlns="http://www.w3.org/1999/xhtml"><p>Hypertension is a common disease in highly industrialized societies, more often perceived as a health problem in adults rather than children. However, epidemiologists are currently paying more attention to the possibility of idiopathic hypertension during childhood. This article compares three classification models (logistic regression, classification trees and MARSplines) in order to determine the best classification model and distinguish the parameters that are most important in the detection of abnormal blood pressure in children. The study group consisted of 1,378 children aged between 7 and 18. After making comparisons between the methods, it was determined that MARSplines is the model that best assigns subjects to classes and can be an alternative in cases when traditional statistical methods cannot be used due to a lack of fulfillment of conditions. For prediction of abnormal blood pressure in this age group, the most important parameters were the heart rate and selected indicators of body proportions.</p></abstract>

Hypertension is a common disease in highly industrialized societies, more often perceived as a health problem in adults rather than children. However, epidemiologists are currently paying more attention to the possibility of idiopathic hypertension during childhood. This article compares three classification models (logistic regression, classification trees and MARSplines) in order to determine the best classification model and distinguish the parameters that are most important in the detection of abnormal blood pressure in children. The study group consisted of 1,378 children aged between 7 and 18. After making comparisons between the methods, it was determined that MARSplines is the model that best assigns subjects to classes and can be an alternative in cases when traditional statistical methods cannot be used due to a lack of fulfillment of conditions. For prediction of abnormal blood pressure in this age group, the most important parameters were the heart rate and selected indicators of body proportions.

Application of Chosen Data Mining Methods in Predicting Abnormal Blood Pressure in Children and Adolescents

Department of Computer Science and Statistics

Studies in Logic, Grammar and Rhetoric

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Hypertension is a common disease in highly industrialized societies, more often perceived as a health problem in adults rather than children. However,...

Application of Chosen Data Mining Methods in Predicting Abnormal Blood Pressure in Children and Adolescents

Online veröffentlicht: 16. März 2019

Seitenbereich: 19 - 28

DOI: https://doi.org/10.2478/slgr-2018-0038

© 2018 Anna Sowińska et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.