Data Mining Approach for the Prediction of Hypertension and its Correlation with Socioeconomic Factors in Mexico: A Case Study

Abstract

According to the Secretary of Health, High Blood Pressure (HBP) has remained among the top ten leading causes of death in Mexico. In recent years, data-driven analysis studies have become a common complement to health research. Therefore, reliable records on this subject are necessary. This paper shows the collection, selection, and integration of a unified database created from different public access sources regarding HBP. We propose a methodology for the identification of correlations between non-medical factors and HBP using data mining techniques such as clustering, we validate them with Pearson Correlation Coefficient. We also used statistical and artificial intelligence models to predict the number of cases of HBP, we evaluated them with Root Mean Square Error and Mean Absolute Percentage Error, the best results were with Convolutional Neural Network Quantile Regression. All in order to generate tools that support the prevention of the future development of hypertension.