Modeling of hypertension risk based on age, body mass index, and psychological factors using local linear estimator

Hypertension, called the silent killer, is a non-infectious disease that is a significant cause of premature death worldwide. Based on WHO, an estimated 46% of adults with hypertension are unaware that they have the condition. Many studies have discussed some factors which can cause hypertension risk such as age, body mass index, and psychology. We need to build a model to estimate hypertension risk based on age, body mass index, and psychological factors. In statistical analysis there are two estimation approaches namely parametric regression and nonparametric regression models. In nonparametric regression, local linear is one of the estimators which has advantages such as it can estimate regression function point-to-point such that it gives results that close to the actual pattern, and no large observations required for assessing the model. In this research, we discuss estimating model of hypertension risk using local linear by including factors of age, body mass index, and psychology. Comparison the proposed model approach and parametric logistic regression approach is also given. The results are that classification accuracies of hypertension risk for factors of age, body mass index, and psychology are 97.62% based on local linear, and 46.4% based on parametric logistic regression. This means that for this case the proposed model approach is better and more appropriate for analyzing the hypertension risk than the parametric logistic regression model approach.